University of Alberta

Socioeconomic Disparities in Eye Care Services and Eye Complications

Among Diabetic Patients in Canada

by

Jongnam Hwang

A thesis submitted to the Faculty of Graduate Studies and Research

in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

in

Health Services and Policy Research

Department of Public Health Sciences

© Jongnam Hwang

Fall 2013

Edmonton, Alberta

Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis

and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is

converted to, or otherwise made available in digital form, the University of Alberta will advise potential

users of the thesis of these terms.

The author reserves all other publication and other rights in association with the copyright in the thesis and,

except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or

otherwise reproduced in any material form whatsoever without the author's prior written permission.

Dedication

This thesis is dedicated to my lovely parents, for their love and support throughout my

life, and for raising me to be the person I am today.

Abstract

Diabetic retinopathy (DR) is a major cause of visual impairment that ultimately impedes

daily activities. Visual impairment caused by DR is manageable if diagnosed early.

Despite comprehensive clinical guidelines, there is underuse of the recommended eye

examinations among patients with diabetes in Canada. This dissertation comprises of four

studies identifying 1) the existence of disparities in eye care services and visual

impairment, 2) the socio-demographic determinants of socioeconomic disparities in eye

care services and visual impairment, and 3) the socioeconomic factors associated with

visual impairment and eye screening services among patients with diabetes.

The results of three separate analyses indicate the presence of 1) socioeconomic

disparities in eye care services at the provincial level and 2) income-related disparities in

visual impairment and eye screening services at the national level. At the provincial level,

income- and material deprivation-related disparities consistently showed a “pro-rich”

pattern, while the social deprivation index indicated a “pro-poor” pattern. In addition,

material deprivation index and place of residence (urban/rural) were important

contributors to the observed income- and material deprivation-related disparities. The

social deprivation-related disparity was explained mainly by social deprivation itself.

At the national level, income-related disparities in eye screening services and preventive

eye screening services revealed a “pro-rich” pattern while the disparity in visual

impairment indicated a “pro-poor” pattern. The main contributor to the observed

disparities in eye screening services was income while the disparity in visual impairment

was predominantly related to age.

In addition, an examination of socioeconomic factors associated with visual impairment

and eye screening services among Canadians living with diabetes provided further

evidence that demographic factors and duration of diabetes were associated with visual

impairment. Regarding eye screening services and preventive eye screening services,

income, patient’s experience in discussing diabetic eye complications with health

professionals and having private insurance covering eye care appointment were

associated with regular eye screening services.

We have contributed new evidence on previously unexplored issues and our work

highlights a need for developing health policy to alleviate the gap in the use of eye

examination across different socioeconomic groups, and for studies providing a better

understanding of the observed disparities.

Acknowledgement

Foremost, I would like to express my deepest gratitude to my supervisor, Dr. Jeffrey

Johnson, who constantly and patiently supported me for my PhD study and research over

the past 4 years. Without his supervision, I could not have successfully completed this

long journey. I would like to thank my co-supervisor and thesis committee members: Dr.

Sarah Bowen, Dr. Chris Rudnisky, Dr. Yutaka Yasui, and Dr. Yukiko Asada, for their

encouragement, insightful comments and rigid questions.

A very special thank you goes to Drs. Soonman Kwon and Jung-Hwa Kim in Korea.

Without their motivation and encouragement, I would not have considered a graduate

career in public health policy research. Drs. Kwon and Kim are the teachers who truly

made a difference in my life. Thank you to Irene Wong, the Statistic Canada analyst at

Research Data Centre at the University of Alberta for her support and advice on data

analyses.

I must also acknowledge my dear friends who incredibly supported and encouraged me

whenever I encountered numerous challenges during my long journey: Robin, Serena,

Calypse, Maria, Bach, Daniala, Fatima, Darren, and Ahmed. Also, thank the Almighty

God for his guidance and wisdom.

Last but not least, I would like to thank the Almighty God for his guidance and wisdom.

Table of Contents

Chapters

   

1. Introduction .................................................................................................................... 1    

2. Socioeconomic Disparities in Eye Care Services among Diabetic Patients in

Alberta, 1995-2009 ........................................................................................................ 15

3. Determinants of Socioeconomic Inequities in the Use of Eye Care Services among

Diabetic Patients in Alberta, 1995-2009 ...................................................................... 54

4. Factor Associated with Eye Screening Services and Visual Impairment in

Canadians Living with Type 2 Diabetes ...................................................................... 88

5. Income-related Disparities in Visual Impairment and Eye Screening Services

in Type 2 Diabetic Patients in Canada ....................................................................... 113

6. Conclusion .................................................................................................................. 141

List of Tables Table numbers prefixed S indicate chapter supplements

Tables 2.1 Total number of diabetic patients who received eye care services by an

ophthalmologist in a 15-year period ..................................................................... 34 2.2 Distribution of eye care services by household income and deprivation indices and

AHS zones in 2009 ................................................................................................ 35 2.3 SES-related Relative Concentration Indices in the use of eye care services, 1995-2005 .............................................................................................................. 36

2.4 Geographic decomposition of SES-related Concentration Indices by AHS zones .............................................................................................................................. 37

S2.1 Alberta physician claims data .............................................................................. 44 S2.2 North zone – Distribution of eye care services by ophthalmologists for Aboriginal and non-Aboriginal people with diabetes, 1995-2009 .......................................... 45

S2.3 North zone – Distribution of eye care services in Aboriginal people with diabetes by income quintile ................................................................................................ 46

S2.4 North zone- Distribution of eye care services in non-Aboriginals by income

quintile .................................................................................................................. 46 S2.5 North zone- Distribution of eye care services in Aboriginals by MDI quintile ... 47

S2.6 North zone- Distribution of eye care services in non- Aboriginals by MDI quintile

............................................................................................................................... 47 S2.7 North zone – Distribution of eye care services by ophthalmologists for urban and rural dwelling people with diabetes, 1995-2009 .................................................. 48 S2.8 North zone- Distribution of eye care services in urban dwellers by income quintile ............................................................................................................................... 49 S2.9 North zone- Distribution of eye care services in rural dwellers by income quintile

.............................................................................................................................. 49

S2.10 North zone- Distribution of eye care services in urban dwellers by MDI quintile .............................................................................................................................. 50

S2.11 North zone- Distribution of eye care services in rural dwellers by MDI quintile

.............................................................................................................................. 50

3.1 Income-related RCI by age, sex, place of residence, Aboriginal status, and

duration of diabetes ............................................................................................. 73 3.2 MDI-related RCI by age, sex, place of residence, Aboriginal status, and duration of diabetes .............................................................................................. 73 3.3 SDI-related RCI by age, sex, place of residence, Aboriginal status, and duration of diabetes ............................................................................................ 74

3.4 Income-related HIs in the use of eye care services among diabetic patients over a 15-year period ........................................................................................... 74

3.5 MDI-related HIs in the use of eye care services among diabetic patients over a 15-year period ........................................................................................... 75

3.6 SDI-related HIs in the use of eye care services among diabetic patients

over a 15-year period .......................................................................................... 75 3.7 Decomposition of income-related RCI, 1995-2009 ............................................. 76

3.8 Decomposition of MDI-related RCI, 1995-2009 ................................................ 77

3.9 Decomposition of SDI-related RCI, 1995-2009 ................................................. 78 S3.1 Alberta physician claims data .............................................................................. 84 S3.2 Distribution of eye care services by household income in 2009 .......................... 85

S3.3 Distribution of eye care services by material deprivation index in 2009 ............. 86

S3.4 Distribution of eye care services by social deprivation index in 2009 ................ 87 4.1 Characteristics of type 2 diabetic patients by eye screening services ............... 104 4.2 Multivariate logistic regression for eye screening services, preventive eye

screening services and visual impairment ......................................................... 106

S4.1 Multivariate logistic regression for preventive eye screening services

in type 2 diabetic patients .................................................................................. 112 5.1 Visual impairment and eye screening services according household income .. 130

5.2 Measurement of disparities for visual impairment and eye screening services . 130

5.3 Decomposition results for visual impairment among type 2 diabetic patients .. 131 5.4. Decomposition results for eye screening services among type 2 diabetic patients ............................................................................................................................ 132

5.5. Decomposition results for preventive eye screening services among type 2 diabetic patients .............................................................................................................. 133

List of Figures

Figures prefixed S indicate chapter supplements

Figures

2.1 Relative Concentration Index in eye care services by ophthalmologists ............. 37

 2.2 The trends of household income-related RCIs in eye care services by ophthalmologists .................................................................................................. 38 2.3 The trends of material deprivation index-related RCIs in eye care services by

ophthalmologists ................................................................................................... 39

 2.4 The trends of social deprivation index-related RCIs in eye care services by ophthalmologists .................................................................................................. 39

 S2.1 Proportion of Aboriginals and non- Aboriginals in North Zone with eye care services by ophthalmologists ................................................................................ 51

 S2.2 Income-related RCI by Aboriginal status in North Zone ...................................... 51

 S2.3 MDI-related RCI by Aboriginal status in North Zone .......................................... 52

 S2.4 Proportion of urban and rural dwellers in North Zone with eye care services by an

ophthalmologist .................................................................................................... 52

 S2.5. Income-related RCI by Urban and rural in North zone ........................................ 53

 S2.6 MDI-related RCI by urban and rural in North Zone ............................................ 53

 3.1 The trends of horizontal inequity indices over a 15-year period, 1995-2009 ....... 79

 5.1 Aggregated contributions to RCI for visual impairment ..................................... 134

 5.2 Aggregated contributions to RCI for eye screening services .............................. 135

 5.3 Aggregated contributions to RCI for preventive eye screening services ............ 136

List of Abbreviations

AAO American Association of Ophthalmology

ADA American Diabetes Association

ADSS Alberta Diabetes Surveillance System

AHS Alberta Health Services

AHW Alberta Health Wellness

CC Concentration curve

CDA Canadian Diabetes Association

CHA Canadian Health Act

CNIB Canadian National Institute for the Blind

DM Diabetic mellitus

DR Diabetes retinopathy

ETDRS Early Treatment of Diabetic Retinopathy Study

FSA Forward Sortation Area

HBM Health Behaviour Model

HI Horizontal inequity index

ICD International Classification of Disease

MDI Material deprivation index

NSPB National Society to Prevent Blindness

NPDR Non-proliferative diabetic retinopathy

OR Odds ratio

PCA Principal component analysis

PDR Proliferative diabetic retinopathy

RCI Relative Concentration Index

SDI Social deprivation index

SES Socioeconomic status

SLCDC Survey on Living with Chronic Disease in Canada

 1

Chapter 1. Introduction

1.1 Diabetes and Diabetic Complications

Diabetes is a common chronic disease responsible for significant morbidity and mortality

in Canada. Approximately 6.2% of Canadians live with diabetes; Alberta’s prevalence is

somewhat lower at 5.7% in 2009 (1-3). The prevalence of diabetes is expected to

continuously increase (4). Several risk factors have been identified as contributing to the

increasing prevalence of diabetes and its complications: the general aging of the

population, and increasing rates of obesity, physical inactivity, and hypertension (4).

The increasing prevalence of diabetes is a substantial concern due to its related long-term

complications such as ischemic heart disease, stroke, renal failure, neuropathy and

retinopathy that directly affect individuals’ daily lives (5-8). These long-term

complications of diabetes are generally grouped into macro- and microvascular

complications. Macrovascular refers to the cardiovascular system, including heart disease,

hypertension and cerebrovascular disease. Macrovascular complications are the leading

cause of morbidity and mortality in people with diabetes. Indeed, patients with diabetes

are two to three times more likely to suffer cardiovascular disease (8-10) and three times

more likely to die of ischemic heart disease than are non-diabetic patients of similar age

and sex (11). Also, people with diabetes have about twice the prevalence of hypertension

and twice the incidence of stroke compared to people without diabetes (3, 6).

Diabetic retinopathy (DR) is a common and specific microvascular complication of

diabetes (12-15). At 20 years after the initial diagnosis, almost every patient with type 1

 2

diabetes, and 60% of people with type 2 diabetes will have some degree of diabetic

retinopathy and, after 30 years, almost all are affected (16). DR is classified into two

main groups: an early stage, non-proliferative diabetic retinopathy (NPDR) and a more

severe stage, proliferative diabetic retinopathy (PDR) (17). Both NPDR and PDR are

associated with worse health outcomes for people with diabetes. NPDR causes vision

impairment through increased intra-retina vascular permeability that ultimately induces

macular edema and variable degrees of intra-retinal capillary closure, causing macular

ischemia. The development of PDR is associated with an increased risk of myocardial

infarction, stroke, diabetic nephropathy, and amputation (17, 18).

DR is the leading cause of blindness in North America (19). It is responsible for about 12%

of all new cases of blindness, affecting more than 8,000 individuals each year in the

United States (13). In Canada, DR causes an estimated 600 new cases of blindness each

year (20). Considering the increasing number of diabetic patients in the Canadian

population, it is anticipated that the prevalence of DR and blindness will continuously

increase (21). It has also been estimated that nearly half a million Canadians currently

have some form of DR and approximately 100,000 Canadians have PDR, diabetic

macular edema or both (19). Visual impairment and blindness caused by DR are

important causes of decreased quality of life (22) and low involvement in the labour force

(16). Vision impairment caused by DR is preventable and made more manageable when

the disease is detected and treated early in its course; however, once vision loss develops,

it is less likely to be recovered (18, 23, 24).

1.2. Screening for Diabetes Retinopathy

 3

Routine screening for DR is recommended in many countries based on intervention and

economic studies. The Canadian Diabetes Association (CDA) recommends an annual

dilated eye examination to prevent potential vision loss (14). Screening for DR is highly

effective, economical, and available under the universal health care system in Canada

(12). Screening guidelines for DR have been also supported by professional

organizations in the United States such as the American Diabetes Association (ADA),

and the American Association of Ophthalmology (AAO) (25). In the U.K., screening for

DR is offered as a national program to reduce the risk of vision loss in people with

diabetes (16). Despite established comprehensive guidelines for routine screening for DR,

underuse of recommended eye screening services is commonly reported among diabetic

patients from different countries, regardless of the type of health care system (24, 26-28).

The gold standard for the detection of DR consists of 30-degree, seven-field, stereoscopic

photography, as developed for the Early Treatment of Diabetic Retinopathy Study

(ETDRS) (25). This has a sensitivity and specificity for the detection of DR that is

superior to direct and indirect ophthalmoscopy by ophthalmologists. Examinations can

also be done using retinal photographs (with or without dilation of the pupil) that are read

by experienced experts in this field (25). In Canada, the current standard of care for

identification of DR is a stereoscopic assessment of the retina through a dilated pupil by

an experienced eye care professional (12, 14). In addition, teleophthalmology programs

have been developed to provide as a modified ETDRS protocol for screening patients

living in remote areas (29-31).

1.3. Definitions of Inequality and Inequity in Health and Health Care

 4

While the terms disparity, inequality and inequity are often used interchangeably in

literature, it is critical to distinguish the concepts and definitions of these three terms.

Disparity is a neutral term used to designate any difference between group and it implies

little implication of social value or justice (32). Either inequality or inequity can be

described as disparity. Inequality is used to explain a difference or variation in health

status or in the distribution of determinants associated with health and health care

between individuals or groups (33). The concept of inequality is not necessarily value

based, so it is more likely a descriptive term without moral judgement (34). An example

of inequality is differences in health outcomes that result from biological endowment,

such as sex. In contrast, health inequity is a difference in health or health care, which is

not justified on the basis of need, as a result of unfairness or injustice in a society (33-35).

Inequity is considered an avoidable and unnecessary difference between different groups

(33, 36). It was suggested that inequality and inequity are not synonymous (34), and these

terms are required to be distinguished in health and health care research (33).

1.4. Inequities in Health and Access to Health Care Services

Accumulating evidence indicates that poor SES is associated with worse health outcomes

and underuse of preventive health care services in the general population (37-42). Lower

income, lower educational attainment, lower wealth, and less privileged occupation at the

individual level have been identified as predictors of inequities (38, 43). In addition, SES

factors contributing to inequities in health and health care encompass not only individual

characteristics, but also community-level characteristics and gradients of socioeconomic

status between individuals and communities (44). These factors can be conceptualized

and measured over an individual’s lifetime (44).

 5

Rigorous research has provided evidence on the relationship between SES and the

prevalence of diabetes and diabetic complications. Dinca-Panaitescu et al. observed a

graded association between SES (based on income and educational level) and prevalence

of diabetes, confirming that social and economic status are important factors influencing

the prevalence of diabetes in Canada (45). Rabi et al. also reported a significant gradient

in prevalence of diabetes across income groups (46). In addition, evidence from rural

Manitoba suggests that this SES-related gradient in prevalence of diabetes has widened

over the last 20 years (47). There is also evidence of higher rates of diabetes-related

amputations in lower income neighbourhoods (47).

1.5. Inequities in Diabetic Eye Complications and Eye Screening Services

Internationally, SES is a key predictor of increased rates of diabetic eye complications

(46, 48-54). In their summary of the literature, Brown et al. found that lower SES, as

measured by individual or household income, education, employment, occupation, or

living in an underprivileged area, is associated with increased risk of microvascular

disease, including ocular complications (44). Rates of ocular complications are also

higher among minority groups, which tend to be of lower SES. A systematic review

conducted in the United States reported that African Americans and Latinos have higher

rates of blindness compared to the general population (52). Within the U.S. population of

visible minority groups, an SES-related gradient has also been found: higher SES groups

have better ocular health outcomes compared to those with lower SES (52). These

findings are a result of multifaceted SES interaction, indicating that race/ethnicity, while

related to ocular complications, is likely attributable in large part to SES (23, 53, 55-57).

 6

Given this evidence, SES appears to be an important determinant of ocular complications

in diabetic populations.

With respect to eye examination services, inequities in the use of eye screening services

are observed among lower SES groups, regardless of the type of health care system (48,

51, 53). Moss et.al found that more education or higher income and having health

insurance that covered eye examinations were predictors of receiving an eye examination

in individuals with younger-onset diabetes (29). Another study found that a patient’s

educational level, health literacy, and language fluency were also identified as barriers to

eye screening services by an ophthalmologist (53). Differences in utilization of

recommended screening services by ethnic/racial groups also suggest a relationship with

SES. For example, the underuse of recommended screening for DR observed in the

Hispanic population could be explained by a phenomenon known as the “accumulative

effect” (52). The accumulative effect refers to fact that systematic societal factors result

in the greater likelihood that racial/language minorities will have lower income and live

in poorer neighbourhoods or remote residential areas (40, 55).

Similar findings have been found in the U.K., a country with a publicly funded health

care system, and in which there are important socioeconomic differentials in the

utilization of preventive medical care, including DR screening, indicating that the rate of

DR screening was significantly lower in patients residing deprived areas (23, 53, 54, 57)

(54, 58, 59). Studies in Australia, Spain, and Hong Kong have also found that the groups

at high risk of underutilization are those who are in poverty or have lower income (60).

Overall, the accumulated evidence indicates a strong association between SES and

utilization of eye screening services: lower SES, including lower income and education,

 7

poverty, and living in a poorer neighbourhood could be considered as predictors of use of

recommended eye examination services in diabetic patients (25).

1.6. Diabetic Eye Complications and Eye Screening Services in Canada

Despite the growing body of literature on inequities in both health outcomes and health

care services in Canada, there is little research focusing specifically on diabetic ocular

complications and eye screening services. Several studies of the general population have

documented that the proportion of people with uncorrected vision problems or blindness

was highest for individuals in the lowest income and educational levels (22). In a recent

Canadian pilot study, Gold et al. reported that the rising cost of visual aids present a

significant barrier to vision rehabilitation service benefits for individuals on low or fixed

incomes, particularly seniors (61). Sit et al., using Canadian National Institute for the

Blind (CNIB) blindness registration data, suggested that, besides age, median household

income is the most common predictors of blindness and poor ocular health outcomes (62).

In another Canadian population-based study, along with age, being female and having a

low income and lower educational attainment were associated with vision problems (22).

In the 2006 study of five provinces and one territory, Sanmartin et al. reported that only

68% of diabetic patients had self-reported receiving the recommended level of eye

screening service (63). This study found that patients aged 18 to 44 were less likely to

have had an eye examination compared with older patients (63). However, due to limited

data collection and lack of information at the individual level, the study was unable to

fully explain the possible causes of the underuse of eye examinations in younger patients.

Another Canadian study that used population-based provincial data on ophthalmic

services for persons with diabetes focused on determining both the direct and indirect

 8

costs of providing ophthalmic, social, and rehabilitative services for persons with diabetic

ocular diseases; however, the study did not measure the association between SES and

diabetic ocular diseases/ophthalmic services including recommended eye screening

services in patients with diabetes (64).

The literature from several countries with different health care systems indicates that

diabetic eye complications and use of eye screening services are consistently worse in

diabetes patients with lower SES. It is unknown whether this same pattern exists among

diabetic patients living in Alberta and Canada. Further research is required to establish

the magnitude of health disparities between those of low, average, and high SES in order

to identify and quantify the level of any inequities.

1.7. Objectives and Program of Research

The objective of this program of research was to measure socioeconomic-related

disparities in visual impairment and eye screening services among patients with diabetes

both provincially (in Alberta) and nationally. We sought to examine whether there are

disparities in visual impairment and eye screening services across different

socioeconomic indicators and to quantify each contribution of socio-demographic factors

to the observed disparities at both levels. We were also interested in assessing whether

any apparent disparities could be defined as either inequities or inequalities. A further

objective was to identify factors associated with visual impairment and eye screening

services among diabetic patients living in Canada.

 9

The first two studies (Chapter 2 and 3) focus on eye care services by ophthalmologists in

Alberta, using three different socioeconomic indicators - household income, material, and

social deprivation indices. These studies utilized data from the Alberta Diabetes

Surveillance System 1995-2009, which is based on Alberta Health administrative

databases. Chapters 4 and 5 focus on visual impairment and eye screening services

among patients with type 2 diabetes using the Survey on Living with Chronic Disease in

Canada – Diabetes Component 2011 (SLCDC –DM 2011), which is a nationally

representative survey dataset.

As a common approach to both data sources (i.e., provincial health care administrative

data and national survey data), the relative concentration index (RCI) and decomposition

analysis were used to address the objectives of our research to better understand

disparities and inequities in visual impairment and eye care services. Using the SLCDC-

DM data to better understand factors associated with visual impairment and eye screening

services, multivariate logistic regression was also used to measure and quantify income-

related disparities among patients with type 2 diabetes.

 10

References

1. Balko SU, Hugel G, Johnson JA. Diabetes Trends in Alberta. Alberta Diabetes Surveillance System Newsletter, 2011. 2. Public Health Agency of Canada. Report from the National Diabetes Surveillance System: Diabetes in Canada. Ottawa: Public Health Agency of Canada; 2009. 3. Public Health Agency of Canada. Diabetes in Canada: Facts ad figures from a public health perspective. Ottawa: Public Health Agency of Canada; 2011. 4. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract. 2010;87(1):4-14. 5. O'Reilly DJ, Xie F, Pullenayegum E, Gerstein HC, Greb J, Blackhouse GK, et al. Estimation of the impact of diabetes-related complications on health utilities for patients with type 2 diabetes in Ontario, Canada. Qua Life Res. 2011;20(6):939-43. 6. Deshpande AD, Harris-Hayes M, Schootman M. Epidemiology of diabetes and diabetes-related complications. Phys Ther. 2008;88(11):1254-64. 7. Kannel WB. Framingham study insights on diabetes and cardiovascular disease. Clinical chemistry. 2011;57(2):338-9. 8. Goff DC, Gerstein HC, Ginsberg HN, Cushman WC, Margolis KL, Byington RP, et al. Prevention of Cardiovascular Disease in Persons with Type 2 Diabetes Mellitus: Current Knowledge and Rationale for the Action to Control Cardiovascular Risk in Diabetes (ACCORD) Trial. Am J Cardiol. 2007;99(12):S4-S20. 9. Kannel WB, McGee DL. Diabetes and cardiovascular disease. The Framingham study. JAMA. 1979;241(19):2035-8. 10. Haffner SM, Lehto S, Rönnemaa T, Pyörälä K, Laakso M. Mortality from Coronary Heart Disease in Subjects with Type 2 Diabetes and in Nondiabetic Subjects with and without Prior Myocardial Infarction. N Eng J Med. 1998;339(4):229-34. 11. Gu K, Cowie CC, Harris MI. Mortality in Adults With and Without Diabetes in a National Cohort of the U.S. Population, 1971–1993. Diabetes care. 1998;21(7):1138- 45. 12. Rudnisky CJ, Tennat MTS, Johnson JA, Balko SU. Chapter 9 : Diabetes and Eye Disease in Alberta. In: Johnson JA,editor Alberta Diabetes Atlas 2011. Edmonton: Institute of Health Economics; 2011. p. 153-74. 13. Public Health Task Force for Vision and Ophthalmology. Determining the information needs for prevention of vision loss strategies in Canada. Public Health Task Force for Vision and Ophthalmology. Can J Ophthalmol. 2000;35(5):255-7. 14. Bond SR AA, Altomare F., Stockl F. Canadian Diabetes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada:

 11

Retinopathy. Can J Diabetes. 2013;37(suppl1):S137-S41. 15. Public Health Agency of Canada. Complications of Diabetes. Ottawa: Public Health Agency of Canada; 2008. 16. Jones S, Edwards RT. Diabetic retinopathy screening: a systematic review of the economic evidence. Diabet Med. 2010;27(3):249-56. 17. American Academy of Ophthalmology. Chapter 5 : Retinal Vascular Disease. Retina and Vitreous Section12 Basic and Clinical Science Course 2011-2012. San Fransico, CA: American Academy of Ophthalmology The Eye M.D. Association; 2011. p. 107- 80. 18. Hooper P, Boucher MC, Cruess A, Dawson KG, Delpero W, Greve M, et al. Canadian Ophthalmological Society Evidence-Based Clinical Practice Guidelines for the Management of Diabetic Retinopathy – Executive Summary. Can J Ophthalmol. 2012;47(2):91-101. 19. Buhrmann R, Houdege W, Beardmore J, Baker G, Lowcock B, Pan I, Bovell A. Chapter 2. The Epidemilogoy of Blindness and Visual Imparment in Canada : Diabetic Retinopathy. Foundations for a Canadian Vision Health Strategy : Towards Preventing Avoidable Blindness and Promoting Vision Health: The National Coalition for Vision Health; 2007. p. 14-5. 20. Szabo SM, Beusterien KM, Pleil AM, Wirostko B, Potter MJ, Tildesley H, et al. Patient Preferences for Diabetic Retinopathy Health States. Invest Ophthalmol Vis Sci. 2010;51(7):3387-94. 21. Aiello LP, Cavallerano J, Bursell SE. Diabetic eye disease. Endocrinol Metab Clin North Am. 1996;25(2):272-90. 22. Perruccio AV, Badley EM, Trope GE. A Canadian population-based study of vision problems: assessing the significance of socioeconomic status. Can J Ophthalmol. 2010;45(5):477-83. 23. Fraser S, Edwards L. Health Equity Audit - Diabetic Retinopathy Screening : Southampton City, Hampshire, Isle of Wight and Portsmouth City PCTs. Southampton City PCT, 2010 July. 24. van Eijk KN, Blom JW, Gussekloo J, Polak BC, Groeneveld Y. Diabetic retinopathy screening in patients with diabetes mellitus in primary care: Incentives and barriers to screening attendance. Diabetes Res Clin Pract. 2011. 25. Garg S, Davis R. Diabetic Retinopathy Screening Update. Clinical Diabetes. 2009;27(4):140-5. 26. Zhang X, Andersen R, Saaddine JB, Beckles GLA, Duenas MR, Lee PP. Measuring Access to Eye Care: A Public Health Perspective. Ophthalmic Epidemiol. 2008;15(6):418-25. 27. Carmen D. Harris LP, Qaiser Mukhtar. Changes in receiving preventive care services

 12

among US adults with diabetes, 1997-2007. Prev Chronic Dis. 2010;7(3):A56. 28. Moss SE, Klein R, Klein BEK. Factors Associated With Having Eye Examinations in Persons With Diabetes. Arch Fam Med. 1995;4(6):529-34. 29. Ng M, Nathoo N, Rudnisky CJ, Tennant MT. Improving access to eye care: teleophthalmology in Alberta, Canada. J Diabetes Sci Technol. 2009;3(2):289-96. 30. Rudnisky CJ, Hinz BJ, Tennant MTS, de Leon AR, Greve MDJ. High-resolution stereoscopic digital fundus photography versus contact lens biomicroscopy for the detection of clinically significant macular edema. Ophthalmology. 2002;109(2):267- 74. 31. Rudnisky CJ, Tennant MTS, Weis E, Ting A, Hinz BJ, Greve MDJ. Web-Based Grading of Compressed Stereoscopic Digital Photography versus Standard Slide Film Photography for the Diagnosis of Diabetic Retinopathy. Ophthalmology. 2007;114(9):1748-544.    32. The Health Disparities Task Group. Reducing Health Disparities- Roles of the Health Sector: Recommended Policy Direnctions and Activities. Ottawa: Public Health Agency of Canada, 2004. 33. Kawachi I, Subramanian SV, Almeida-Filho N. A glossary for health inequalities. J Epidemiol Community Health. 2002;56(9):647-52. 34. Braveman P, Gruskin S. Defining equity in health. J Epidemiol Community Health. 2003;57(4):254-8. 35. Braveman PA, Cubbin C, Egerter S, Chideya S, Marchi KS, Metzler M, et al. Socioeconomic status in health research: one size does not fit all. JAMA. 2005;294(22):2879-88. 36. Dahlgren G, Whitehead M. Policies and strategies to promote social equality in health. Stockholm: Institute of Future Studies, 1991. 37. Raphael D. Health inequalities in Canada: current discourses and implications for public health action. Critical Public Health. 2000;10(2):193-216. 38. Morris S, Sutton M, Gravelle H. Inequity and inequality in the use of health care in England: an empirical investigation. Social Science & Medicine. 2005;60(6):1251-66. 39. Adda J, Chandola T, Marmot M. Socio-economic status and health: causality and pathways. J Econometrics. 2003;112(1):57-63. 40. House JS, Williams DR. Understanding and reducing socioeconomic and racial/ethnic disparities in health. In: Smedley BS, SL., editor. Promoting health: intervention strategies from social and behavioral research. Washington, DC: National Academy Press; 2000. p. 81-124. 41. Warnecke RB, Oh A, Breen N, Gehlert S, Paskett E, Tucker KL, et al. Approaching health disparities from a population perspective: the National Institutes of Health

 13

Centers for Population Health and Health Disparities. Am J Public Health. 2008;98(9):1608-15. 42. Lynch J, Smith GD, Harper S, Hillemeire M, Ross N, Kaplan GA. Explaining the social gradient in coronary heart disease: comparing relative and absolute risk approaches. J Epidemiol Community Health. 2006;60(5):436-41. 43. Lynch J, Smith GD, Harper S, Hillemeire M, Ross N, Kaplan GA, Wolfson M. Is Income Inequality a Determinant of Population Health? MILBAN Q. 2004;82(1):5-99. 44. Brown AF, Ettner SL, Piette J, Weinberger M, Gregg E, Shapiro MF, et al. Socioeconomic Position and Health among Persons with Diabetes Mellitus: A Conceptual Framework and Review of the Literature. Epidemiol Rev. 2004;26(1):63- 77. 45. Dinca-Panaitescu S, Dinca-Panaitescu M, Bryant T, Daiski I, Pilkington B, Raphael D. Diabetes prevalence and income: Results of the Canadian Community Health Survey. Health Policy. 2011;99(2):116-23. 46. Rabi DM, Edwards AL, Southern DA, Svenson LW, Sargious PM, Norton P, et al. Association of socio-economic status with diabetes prevalence and utilization of diabetes care services. BMC health services research. 2006;6:124. 47. Martens PJ, Brownell M, Au W, MacWilliam L, Prior H, Schultz J, Guenette W, Elliott L, Buchan S, Anderson M, Caetano P, Metge C, Santons R, Serwonka K. Chapter 5 : Adult Health. Health Inequities in Manitoba : Is the Socioeconomic Gap Widening or Narrowing Over Time? Winnipeg, MB: Manitoba Centre for Health Policy; September 2010. p. 81-114. 48. Chaturvedi N, Stephenson J, Fuller J, Complications TEI. The Relationship Between Socioeconomic Status and Diabetes Control and Complications in the EURODIAB IDDM Complications Study. Diabetes care. 1996;19(5):423-30. 49. Fang R, Kmetic A, Millar J, Drasic L. Disparities in Chronic Disease Among Canada's Low-Income Populations. Prev Chronic Dis. 2009;6(4):A116. 50. Garcia AA, Benavides-Vaello S. Vulnerable populations with diabetes mellitus. Nurs Clin N Am. 2006;41(4):605-23, vii. 51. Gulliford MC, Dodhia H, Chamley M, McCormick K, Mohamed M, Naithani S, et al. Socio-economic and ethnic inequalities in diabetes retinal screening. Diabet Med. 2010;27(3):282-8. 52. Peek M, Cargill A, Huang E. Diabetes Health Disparities. Med Care Res Rev. 2007;64(5 suppl):101S-56S. 53. Millett C, Dodhia H. Diabetes retinopathy screening: audit of equity in participation and selected outcomes in South East London. J Med Screen. 2006;13(3):152-5. 54. Kliner M, Fell G, Gibbons C, Dhothar M, Mookhtiar M, Cassels-Brown A. Diabetic retinopathy equity profile in a multi-ethnic, deprived population in Northern England.

 14

Eye (Lond). 2012. 55. Williams DR, Mohammed SA, Leavell J, Collins C. Race, socioeconomic status, and health: complexities, ongoing challenges, and research opportunities. Ann NY Acade Sci. 2010;1186:69-101. 56. D'Anna L, Ponce N, Siegel J. Racial and ethnic health disparities: evidence of discrimination's effects across the SEP spectrum. Ethnicity & health. 2010;15(2):121- 43. 57. Scanlon PH, Carter SC, Foy C, Husband RF, Abbas J, Bachmann MO. Diabetic retinopathy and socioeconomic deprivation in Gloucestershire. J Med Screen. 2008;15(3):118-21. 58. Nsiah-Kumi P, Ortmeier SR, Brown AE. Disparities in diabetic retinopathy screening and disease for racial and ethnic minority populations--a literature review. J Natl Med Assoc. 2009;101(5):430-7. 59. Hippisley-Cox J, O'Hanlon S, Coupland C. Association of deprivation, ethnicity, and sex with quality indicators for diabetes: population based survey of 53,000 patients in primary care. BMJ. 2004;329(7477):1267-9. 60. Bragge P, Gruen RL, Chau M, Forbes A, Taylor HR. Screening for presence or absence of diabetic retinopathy: a meta-analysis. Arch Ophthalmol. 2011;129(4):435- 44. 61. Gold D, Zuvela B, Hodge WG. Perspectives on low vision service in Canada: a pilot study. Can J Ophthalmol. 2006;41(3):348-54. 62. Sit AJ, Chipman M, Trope GE. Blindness registrations and socioeconomic factors in Canada: an ecologic study. Ophthalmic epidemiol. 2004;11(3):199-211. 63. Sanmartin C, Gillmore J. Diabetes care in Canada: Results from selected provinces and territory, 2005. Ottawa: Statistics Canada, 2006. 64. Smith AF. The economic impact of ophthalmic services for persons with diabetes in the Canadian Province of Nova Scotia: 1993-1996. Ophthalmic epidemiol. 2001;8(1):13-255  

 

15

Chapter 2. Socioeconomic Disparities in Eye Care Services among Diabetic Patients in Alberta, 1995-2009

Abstract

Diabetic retinopathy (DR) is a leading cause of blindness in Canada. Despite established

comprehensive guidelines for routine DR screening, the rate of provision eye care

services in diabetic patients in Alberta has decreased over the past years. Lower

socioeconomic status (SES) has been identified as a factor associated with

disproportionate use of health care services. While there is increasing evidence for

unequal use of health care services by individuals with lower SES over the past decade,

there is limited evidence regarding the relationship between lower SES and the use of eye

care services by diabetic patients in Canada. The aim of this study was to examine

socioeconomic status (SES)-related disparities in the use of ophthalmologists’ services

for diabetic patients in Alberta. We used data from the Alberta Diabetes Surveillance

System (ADSS), including visits to ophthalmologists over a 15-year period (1995-2009).

Socioeconomic disparities in the use of eye care services were assessed using the Relative

Concentration Index (RCI), which ranges from -1 to +1. We used three different SES

indicators: median household income and the Canadian material and social deprivation

indices (MDI and SDI). We found that socioeconomic disparities in use of eye care

services exist in diabetic patients, although the magnitudes of the disparities were small

and have steadily decreased over time. Income- and MDI-related RCI indicated a “pro-

rich” direction, suggesting that individuals with more income or less material deprivation

were more likely to use eye care services. For SDI-related RCI, socially deprived

individuals were more likely to use eye care services compared to less socially deprived

groups (RCI: 1995:- 0.066; 2002:-0.036; 2009:-0.028). Socioeconomic disparities varied

by Alberta Health Services (AHS) zones, and it was noteworthy that the pattern of change

 

16

in the different RCI varied also by sub-population, particularly in the non-Aboriginal

population in the North zone. Our findings suggest policy development at the provincial

level may be considered to alleviate the observed disparities in the use of eye care

services among diabetic, particularly as it relates to geographical differences across AHS

zones.

 

17

2.1. Introduction

Diabetic retinopathy (DR) is a common and serious life-threatening complication of

diabetes (1-3). Nearly 60~80% of people with type 2 diabetes will have some degree of

diabetic retinopathy 15 to 20 years after their initial diagnosis, and almost all are affected

after 30 years (4, 5). DR is classified into two stages: an early stage, termed non-

proliferative diabetic retinopathy (NPDR), and a more severe stage, termed proliferative

diabetic retinopathy (PDR) (6). Both NPDR and PDR are associated with worse health

outcomes. NPDR and PDR cause vision impairment through increased intra-retinal

vascular permeability that ultimately results in macular edema and varying degrees of

intra-retinal capillary closure, which causes retinal ischemia and neovascularization (6).

In addition to a higher risk of various complications (6), DR is the primary cause of

blindness in North America (2). It is responsible for about 12% of all new cases of

blindness, affecting more than 8,000 individuals each year in the U.S (7). In Canada, DR

causes an estimated 600 new cases of blindness each year (8). One Canadian study has

estimated that DR prevalence could be as high as 40% among patients with diabetes (7).

Given the increasing number of patients with diabetes in Canada, the prevalence of DR is

also expected to continuously increase (9). The visual impairment and blindness caused

by DR are important causes of decreased quality of life (10) and a major impediment to

labour force participation (4). Vision impairment caused by DR can be preventable and

more manageable when the disease is detected and treated early in its course; however,

once vision loss develops, it is less likely to be recovered (11-13).

 

18

Despite established comprehensive guidelines for routine screening, underuse of

recommended eye care services has been observed in patients with diabetes (12, 14-16).

Provincial diabetes surveillance data on eye examinations during the period of 2001-2006

indicates that only 48% of diabetic patients in Alberta received an eye examination by an

ophthalmologist within 3 years of their diagnosis of diabetes (13). These lower rates are

also observed in other parts of Canada. In Ontario, only 19% of patients with diabetes

have undergone a follow-up eye examination after their initial eye exam (17).

Lower socioeconomic status (SES) has been associated with disproportionate use of eye

care services. Moss et. al found that more education, higher income, and having health

insurance that covers eye examination were predictors of receiving an eye examination in

younger diabetic patients (16). In the U.K, which has a publicly funded healthcare system,

the rate of DR screening was significantly lower in patients residing in deprived areas

(18-20). The evidence for SES-related barriers to access to health care services (21-24)

and a growing number of lower SES individuals during the past decade (25) directs our

attention to the potential effects of lower SES on the use of eye care services among those

living with diabetes. It has also been suggested that micro-level investigations focusing

on specific service categories are needed to provide information that cannot be achieved

by a macro-level study of health care equity (26).

Therefore, our study was designed to look for SES-related disparities in the use of eye

care services in patients with diabetes living in Alberta over a 15-year period (1995-2009).

In addition, we examined the geographic decomposition of socioeconomic disparities in

 

19

eye care services among diabetic patients across different Alberta Health Services (AHS)

zones.

2.2. Methods

Our study used data from the Alberta Diabetes Surveillance System (ADSS). The ADSS

was created to disseminate information on the incidence, prevalence, and mortality, and

complications of diabetes in Alberta, Canada (27, 28). The ADSS datasets are created by

linking and de-identifying data from several Alberta Health (AH) databases including

discharge abstract databases, the physician claims database, the ambulatory care

classification system, and the population registry. All diabetic patients, diagnosed by

physicians and covered by the provincial health care plan from 1995-2009 were included

in the ADSS datasets. The case definition to identify diabetic patients requires that an

individual must have one hospitalization with an International Classification of Disease,

9th revision (ICD-9) code of 250, selected from all available diagnostic codes from the

hospital discharge abstract for the years 1995-2001, or equivalent ICD-10 codes (E10-

E14) for the years after 2001-2002, or two physician claims with an ICD-9 code of 250

within 2 years, selected from any of the three available diagnostic codes from the

physician claims database (27). Women with gestational diabetes were excluded (27).

We included only diabetic patients over age 20 years who were classified as adult

diabetic patients by the ADSS case definition, and individuals whose postal code were

linkable to a specific neighbourhood.

We used three different SES indicators: median household income and the Canadian

material and social deprivation indices (MDI and SDI) (28), linked to the ADSS datasets

 

20

at Alberta’s 70 sub-region levels. Median household income was obtained from 2006

Canadian census data. The Canadian MDI and SDI, also created based on 2006 Canadian

census data, were developed to measure multiple dimensions of SES and are widely used

as a measurement of SES (29). MDI is a composition of educational attainment,

employment status and income level. SDI includes the prevalence of single-parent

families as well as the proportion of people living alone, and those who are separated,

divorced or widowed.

The use of eye care services provided by an ophthalmologist after the initial diagnosis of

diabetes was measured by all contacts with ophthalmologists in each year from 1995-

2009 (13). This includes any service or procedure performed by an ophthalmologist and

claimed for reimbursement by Alberta Health [Table S2.1]. Until April 2007, optometrist

services were not fully covered as an insured benefit and therefore not included in the

physician claims database.

To examine SES-related disparities in eye care services by ophthalmologists, we used the

Concentration Curve (CC), and the Relative Concentration Index (RCI). The CC and RCI

were proposed by Kakwani and Wagstaff (30, 31) and are commonly used as standard

measurement tools for socioeconomic-related inequality in the field of health economics

and policy (32, 33). To present the CC, we plotted the cumulative percentage of the

outcome variable (i.e., eye care services by ophthalmologists) on the y-axis, against the

cumulative distribution of individuals by quintiles of SES indicator (i.e., each of

household income, MDI and SDI) plotted on the x-axis. This curve allowed us to measure

the distribution of eye care services by aggregating across individuals for SES-related

 

21

quintiles. The line of equity (i.e., a 45-degree line) would represent an equal distribution

of eye care services (i.e., 20%) across each of the five SES-related quintiles.

To quantify the magnitude of disparities in eye care services by ophthalmologists, we

used the RCI, which is directly related to the CC (29,30). The RCI is defined as twice the

area between the CC and the line of equity. The RCI is typically bounded between -1 and

1. When the outcome variable is dichotomous, the RCI bounds µ -1 and 1- µ,

respectively, where µ is mean of the health care variable (34). The RCIs in the use of eye

screening services among diabetic patients over the 15-year period were calculated based

on the following equation (33, 35) :

𝐶 =2𝑁µμ

ℎ!𝑟!   − 1 −  1𝑁

!

!!!

Where ℎ! is the outcome, eye care services by ophthalmologists, µ is the mean of the

outcome, and 𝑟!=i/N is the fractional rank of individuals I in the median SES (i.e.,

household income, and MDI or SDI) distribution, with i=1 for the lowest and i=N for the

highest. In the case in which there is no SES-related disparity (i.e., the CC equals the line

of equity), the RCI is calculated as zero. By convention, the RCI takes a negative value

when the curve lies above the line of equity, indicating disproportionate concentration of

eye care services among the lowest quintiles (i.e., “pro-poor”), and a positive value when

it lies below the line of equality (i.e., “pro-rich”). The greater value of the RCI means a

greater degree of concentration in a negative or positive direction.

 

22

In addition, we examined geographic decomposition of socioeconomic disparities in eye

care services across five different AHS zones (i.e., South, Calgary, Central, Edmonton,

and North) in order to understand “between zone” and “within zone” contributions to

income-and deprivation-related disparities. We compared the between and within zone

variations in RCI using a method of decomposition of the RCI, based on the following

equation (37, 38).  

C = 𝐶! + 𝛼!𝐶! + 𝑅!

Where C is the calculated RCI, 𝐶! is the between-zone RCI, 𝛼! is the product of the each

area’s population share and its share of eye care services and 𝐶! is the RCI of each AHS

zone, and R is a re-ranking term. 𝐶! is computed by assigning all individuals in a given

zone with mean of value of the use of eye care services in that zone, rank ordering AHS

zones by their mean per capita income, MDI or SDI, and computing the corresponding

RCI for the use of eye care services. 𝐶! indicates the extent of socioeconomic-related

disparity in the use of eye care services within each zone. The weighted sum of these N

CIs captures the fact that within the zone the poor may systematically have lesser or

greater use of eye care services.

We also stratified zone-specific disparities based on Aboriginal status and urban or rural

place of residence. Urban or rural residence was classified by the second digit of the

postal code for home address in the Stakeholder Registry (36). The Forward Sortation

Area (FSA) with the digit ‘0’ was defined as rural place of residence and all other digits

in the FSA were defined as urban place of residence. Aboriginal status was identified

from the Alberta Health and Wellness (AHW) registry, which allows us to identify Status

 

23

Aboriginals (36). We used STATA 12 for Microsoft Windows for descriptive analyses

and to calculate the CC and the RCI, as well as the geographic decomposition.

2.3. Results

Eye Care Services by Ophthalmologists

Descriptive statistics for the number of patients who had an eye care examination by an

ophthalmologist over a 15-year period are presented in Table 2.1. The average annual rate

of eye care services in 15 years was 31.7%, and during the study period, the rate of eye

care services did not change dramatically. The highest rate of eye care services by

ophthalmologists was in 2001, when approximately 40,000 (34.1%) of 116,235 diabetic

individuals were assessed. The lowest rate of eye examination was in 1995, when only

22,647 (28.4%) of 79,743 diabetic individuals received eye care services.

Table 2.2 shows the distribution of eye care services by ophthalmologists across

household income and deprivation indices quintiles in 2009, and the associated RCI, for

the AHS zones and the province as a whole. Generally, the rates of eye care services

provided by ophthalmologists for those living with higher income or with less material

deprivation were higher than those for their counterparts. In contrast, diabetic patients

from more socially deprived neighbourhoods used more eye care services during the

same period.

Changes in socioeconomic related-disparities

 

24

Income-related disparities in the use of eye care services by ophthalmologists were

observed over the 15 year-period (Table 2.3; Figure 2.1). The disparity has been in favour

of higher income groups (i.e., pro-rich), suggesting that patients residing in higher

income neighbourhoods were more likely to receive eye care services. The income-

related RCI was relatively small in magnitude, and has not substantially changed over

time; the RCI was 0.0309 in 1995, 0.0273 in 2002, and 0.0214 in 2009 (Figure 2.1).

Both MDI and SDI-related RCIs got attenuated over the period between 1995 and 2009;

in both cases moving toward the line of equity (Table 2.3; Figure 2.1). It is noteworthy

that while the MDI-related RCIs continuously showed “pro-rich” during the given period,

the social deprivation-related RCI indicated “pro-poor” disparity, which suggests that

individuals with diabetes residing in more socially deprived areas tended to use more eye

care services compared with those residing in less socially deprived areas.

Socioeconomic disparities by Alberta Health Services (AHS) Zones

We observed considerable variation in the SES-related disparities in eye care services

between AHS zones; the pattern of variation depended on the particular SES indicator.

For example, in 2009 (Table 2.2), the North Zone had the greatest pro-rich disparity

based on household income, followed by the South, then Central zones, with relatively

little disparity in Edmonton or Calgary. While the overall disparity associated with MDI

was larger than household income, between-zone variation was less obvious, although the

Central and North displayed greater pro-rich disparities than Edmonton, South or Central

zones (Table 2.2). On the other hand, for the SDI in 2009, the North zone displayed a

small pro-rich disparity while all other zones displayed pro-poor disparities (Table 2.2).

 

25

Based on the geographic decomposition of the RCIs (Table 2.4) in the period 1995-2009,

the “between zone” contributions to income-related RCIs were always greater than the

“within zone” contribution. Similarly the MDI-related RCIs have decreased over time,

but the variations among zones were always greater than variations within zones. In

contrast to MDI-related RCIs, the contributions of “between zone” and “within zone” in

SDI-related RCIs has changed over time, with relatively little “between zone”

contribution and much more in the residual.

Plotting the trends in income-, MDI and SDI-related disparities in eye care services by

AHS zones presents an interesting story (Figures 2.2-2.4). In terms of household income-

related RCI, South zone has consistently had a larger magnitude of income-related RCIs

among the five AHS zones. Calgary and Edmonton zones have consistently shown little

household income-related disparities, while Central and North had continual increases in

income-related RCIs over the 15-year period. In particular, a dramatic escalation in RCIs

between 2004 and 2009 was observed in North zone (Figure 2.2).

For MDI-related RCIs, both Central and North zones had widening “pro-rich” disparities

over a 15-year period, dramatically so for the North zone after 2004 (Figure 2.3). In

contrast, the MDI-related RCIs for the remaining zones showed a constant decrease, in all

cases moving toward the line of equity. With respect to SDI-related RCIs (Figure 2.4),

North zone moved in a direction of “pro-rich” disparity. That is, individuals residing in

less socially deprived neighbourhood were more likely to use eye care services. The

remaining zones showed constant “pro-poor” disparities in the given period. Interestingly,

 

26

there appeared to be a sharp widening in “pro-poor” disparity in Central zone between

1997 and 2000.

Because of the dramatic change in income- and MDI-related RCIs for the North zone, we

further explored the trends in this zone, stratified by Aboriginal status and urban or rural

residence. When stratified by Aboriginal status [Table S2.2-S2.6; Figure S2.1-S2.3], we

see similar rates of eye care services over time. However, while the income-related RCI

for the Aboriginal population remained relatively stable with little disparity over time, it

increased substantially and almost linearly over time in the non-Aboriginal population in

the North zone. MDI-related disparities increased in the non-Aboriginal population of the

North zone after 2004, while in the Aboriginal population, MDI-related disparities (which

were originally pro-rich) shifted to pro-poor after 2001. Thus, in recent years, it appears

that Aboriginal people living in more materially deprived areas of the North zone were

receiving more eye care services by ophthalmologists than Aboriginals living in less

materially deprived areas of the North, while the opposite was true for the non-Aboriginal

population. There appeared to be no difference in the widening pro-rich income-related

disparities in the North zone between urban and rural dwellers, but in relation to the MDI

related disparities, the greater increase of pro-rich disparity was primarily observed in the

urban-dwelling population in the North zone [Table S2.7-S2.11; Figure S2.4-S2.6].

2.4. Discussion

Using provincial diabetes surveillance data, we demonstrated that socioeconomic

disparities in the use of eye care services provided by ophthalmologists to diabetic

patients have existed in Alberta over the past 15 years. In particular, individuals living in

 

27

higher income neighborhoods and less materially deprived areas were more likely to

receive eye care services, while, in contrast, diabetic patients living in more socially

deprived areas were more likely to use the services. In general, however, the magnitudes

of the disparities were relatively small. Furthermore, the MDI- and SDI-related disparities

decreased over time. Interestingly, the trends in change over time of SES-related

disparities varied considerably by zone, and depended on the particular SES indicator, as

well as characteristics of sub-populations, namely Aboriginal status and urban versus

rural residence.

The use of eye care services among diabetic patients may be influenced by a number of

factors. Some of these factors include medical conditions and health behaviours at the

individual level, but SES is also important. Despite Canada’s universal health care system,

which aims to minimize financial barriers to medically necessary healthcare, a vast

literature has revealed that the use of health care services differs by socioeconomic

condition (21, 39-41). For example, people with lower SES tend to use fewer specialist

preventive services, but are more likely to see family physicians and use hospital services

(26, 42, 43).

Given our study design, the decrease in disparities observed in our study may be related

to increasing numbers of patients with a longer duration of diabetes in each of the years

of follow-up. A longer duration of diabetes may foster patient understanding of the use of

recommended care services and the rationale for the recommended care provided by their

family doctor may lead patients to more active involvement in diabetes care in the

primary care setting (44). In Ontario, the overall rate of undergoing a follow-up

 

28

examination was only 19% within one year after the initial eye exam. However, the

overall rate of a follow-up exam climbed to 55% at two years and 84% at four years after

the initial assessment (17). Given the association between the duration of diabetes and

compliance with recommended care, our finding could be interpreted by the fact that

patients residing in less materially or more socially deprived areas receive more eye care

services as their duration of diabetes increases.  

With respect to income- and MDI-related disparities, our results are consistent with

previous findings that higher SES is associated with increased the use of specialist

services. Our findings with respect to trends of disparities in eye care services were

broadly similar to those from a previous Canadian study (26) that also used RCIs to

measure health care disparities in both family physician and specialist visits across

Canadian provinces, demonstrating a “pro-rich” disparity for specialist visits. In

particular, these authors reported a “pro-rich” inequity in specialist visits in Alberta, even

after adjusting for number of specialist visits (26). It seems reasonable, then, to conclude

that our findings show that lower neighbourhood income and more materially deprived

residential areas are important conditions associated with the use of eye care services by

ophthalmologists in Alberta. Our finding also suggest that the underuse of eye care

services by people with lower economic status could cause further widen the gap in eye

health outcomes, although the magnitude of disparities are relatively small (43). This

potential implication could strengthen our concern that Canada’s universal health care

system may reduce economic barrier to eye care services by eye care services by

ophthalmologists, but does not completely eliminate it.

 

29

Unlike income or MDI-related disparities, SDI-related RCIs in the use of eye care

services showed a “pro-poor” trend in disparities; that is, diabetic patients residing in

more socially deprived area received more eye care services. The explanation for this

opposing trend is not clear. One might argue that given the higher prevalence of diabetes

in lower SES populations, there would also be greater prevalence of diabetic

complications such as eye disease, and therefore more eye care services would be utilized.

That explanation, however, is contradicted by the pro-rich disparities based on income

and MDI. The difference, then, must be based on the different constructs of social

deprivation relative to material deprivation, or the statistical construction of these two

indices (28).

One Israeli study has attempted to explain the higher use of health care services in lower

social condition groups. That study suggest that this finding could be explained by the

fact that unemployed people, in particular retired individuals and individuals with co-

morbid conditions, would be more likely to use specialist services (45). This suggests that,

once patients leave the workforce, and have worse health outcomes, they would focus

more time and attention on their health care, which is reflected by more services being

performed. This explanation may help us understand the direction of disparity related to

social deprivation RCIs found in our study.

Alternatively, the fact that there are more socially deprived areas in urban settings, where

more ophthalmology services are available, could help explain the “pro-poor” direction of

SDI-related RCIs. In our descriptive analysis of the distribution of SDI ranks across the

province, lower SDI ranks were more concentrated in Edmonton and Calgary zones, but

 

30

more than 80% of ophthalmologists offer eye care services in these two zones (data not

shown). It is therefore plausible that a large number of individuals residing in more

socially deprived neighbourhood in urban areas could easily access eye care services, and

this may result in pro-poor disparities related to social deprivation RCIs. As a result,

social deprivation may not be a strong predictor of the use of health care services. Due to

the complexity of social capital, which links closely to the SDI, it has been suggested that

not all components of social capital are related to use of health care services (46).

One possible explanation for the different direction of SDI-related RCI is the

independency between material and social deprivation indices. Material and social

deprivation index were constructed with principal component analysis (PCA), with

application of a varimax rotation in order to improve readability and make these two

factors independent (29). The scoring for using PCA forces the MDI and SDI to be

uncorrelated. Still, the lack of correlation does not obviously explain the opposite

relationship of MDI- and SDI-related RCI over time in the province as a whole. To

understand the factors that might explain the observed “pro-poor” trend in SDI-related

disparities, further research is required.

Uneven distribution of health care resource across the population has been a long-time

public health concern in Canada (47, 48). Our results indicate regional disparities in the

use of eye care services between AHS zones. For both income- and MDI-related RCIs,

disparities in the use of eye care services among AHS zones were consistently larger than

those within each zone. This could simply indicate an unequal distribution of eye care

services resources across Alberta. Approximately 85% of ophthalmologists provide their

 

31

services in urban areas such as Edmonton or Calgary as most Albertans live in these areas

(49). This concentrated distribution of ophthalmologists in Edmonton and Calgary zones

may be associated with regional disparities as diabetic patients living in those areas could

access their eye care services more easily than those living in other AHS zones.

Geographical disparities in the use of eye care services related to SDI-related RCIs were

also observed during the same period, although the contribution of geography to between

zone disparities became smaller than within zone disparities. Three AHS zones--- South,

Calgary, and Edmonton---showed a relatively small magnitude of pro-poor disparities

and moved closer to the line of equity.

The most notable finding was the change in income- and MDI-related disparities in the

North zone, with increasingly pro-rich disparities after 2004. Given the rapid change in

these disparities, it seems unlikely to be driven by specific demographic characteristics of

the population. The pattern of change in disparities was not associated with Aboriginal

status, but seemed to be more apparent in the non-Aboriginal urban-dwelling population

in the North. One explanation for the rapid change in SES-related disparities in the North

may be related to changes in policies or delivery of eye care services. For instance, the

introduction of tele-ophthalmology services in Northern Alberta communities began in

1998 to provide recommended eye screening services for DR and eye care services: this

was intended to address geographical barriers in access to ophthalmologist services. This

introduction of the comprehensive tele-ophthalmology program may have inadvertently

contributed to income- and MDI-related disparities in Northern Alberta communities (50).

Limitations

 

32

While the CC and RCI are widely used for quantifying the magnitude of socioeconomic

inequality in health and health care, their use has well-known limitations that should

make us cautious when interpreting the results of our study. Firstly, an obvious limitation

of using the RCI as a measure of SES-related disparities is that the RCI does not take into

account other individual-level factors associated with the use of eye care services such as

clinical need, sex, age, and immigrant status (51). To mitigate this limitation, we used the

Canadian deprivation index, which reflects various socioeconomic conditions including

educational attainment, employment and social network.

Secondly, using income and the Canadian deprivation index at the sub-region level may

limit our ability to understand income and deprivation at the individual level. Thus, SES

measures at sub-region level could overlook variations in income and deprivation index

among individuals in the same neighbourhood. Nevertheless, neighbourhood is an

important factor affecting health outcomes and healthcare use in Canada as

neighbourhoods are segregated by socioeconomic status. It, therefore, becomes crucial to

understand the characteristic of neighbourhoods in tackling disparities in health care (52).

Thirdly, using provincial health care administrative data dose not allow us to assess the

clinical need and appropriateness of the eye care services received or not received. In

addition to this limitation in the classification of eye care services, provincial

administrative database only included ophthalmology services, so patients who used

optometry services were excluded from our analyses. As of 2007, eye examinations by

optometrists for patients with diabetes have been covered by the provincial healthcare

 

33

plan (13). Further examination of disparities in eye screening services should include

information on services by both ophthalmologists and optometrists.

2.5. Conclusion

Our study suggests that household income-, MDI and SDI-related disparities in the use of

eye care services were observed among diabetic patients living in Alberta. These

disparities are not large, however, and trends over time suggest that they have been

decreasing. However, while there appears to be very little SES-related disparities in the

province overall, patterns of disparities in the use of eye care services among diabetic

patients varied across AHS zones. Our findings suggest policy development at the

provincial level may help to alleviate the observed disparities in eye care services among

patients living with diabetes, particularly in the non-metro areas of the province.

Nonetheless, further study is necessary to better understand the patterns and interpretation

of socioeconomic disparities over time and geographical difference across AHS zones.

 

 

 

 

 

 

 

 

 

 

34

Tables and Figures

Table 2.1. Total number of diabetic patients who received eye care services by an ophthalmologist in a 15-year period

Year  Eye  care  services  

Total   Rate  of  Service  (%)  Yes   NO  

1995   22,647   57,096   79,743   28.4  

1996   25,265   58,110   83,375   30.3  

1997   28,521   59,153   87,674   32.5  

1998   30,909   61,924   92,833   33.3  

1999   33,800   66,057   99,857   33.9  

2000   35,858   71,813   107,671   33.3  

2001   39,650   76,585   116,235   34.1  

2002   42,286   83,143   125,429   33.7  

2003   44,686   89,333   134,019   33.3  

2004   47,351   96,402   143,753   32.9  

2005   49,004   104,908   153,912   31.8  

2006   50,549   114,247   164,796   30.7  

2007   51,694   123,979   175,673   29.4  

2008   53,814   132,154   185,968   28.9  

2009   57,237   141,323   198,560   28.8  

 

 

 

 

 

 

 

 

 

 

 

35

Table 2.2. Distribution of eye care services by household income and deprivation indices and AHS zones in 2009

Household  Incomea   South   Calgary   Central   Edmonton   North   Total  

1st  (Lower)   30.2%   32.9%   18.9%   25.5%   21.0%   26.3%  2nd   37.3%   32.8%   15.9%   28.2%   16.6%   27.7%  3rd   29.2%   33.3%   20.1%   25.2%   30.6%   30.8%  4th   52.9%   33.3%   26.9%   26.7%   35.1%   29.6%  

5th(Higher)   35.7%   29.8%   18.7%   27.8%   24.4%   29.9%  RCI   0.0989   -­‐0.0120   0.0637   0.0071   0.1164   0.0214  

Material  Deprivationa   South   Calgary   Central   Edmonton   North   Total  

1st  (More)   26.5%   26.0%   16.5%   24.9%   23.7%   26.0%  2nd   37.5%   28.6%   21.7%   28.0%   17.6%   28.6%  3rd   50.0%   28.3%   18.3%   23.5%   29.4%   28.3%  4th   35.7%   28.4%   21.0%   25.8%   35.1%   28.4%  

5th  (Less)   37.3%   33.2%   25.1%   31.7%   23.3%   33.2%  RCI   0.0309   0.0238   0.0832   0.0339   0.0796   0.0427  Social  

Deprivationa   South   Calgary   Central   Edmonton   North   Total  

1st  (More)   37.3%   34.6%   28.1%   27.1%   20.7%   30.5%  2nd   23.1%   35.0%   17.1%   26.6%   41.7%   32.7%  3rd   52.9%   31.5%   18.9%   27.8%   16.3%   27.1%  4th   31.7%   31.7%   21.1%   26.3%   20.9%   25.4%  

5th  (Less)   34.8%   29.2%   18.7%   24.1%   31.2%   28.4%  RCI   -­‐0.0056   -­‐0.0335   -­‐0.0716   -­‐0.0159   0.0167   -­‐0.0278  

 

a)  Higher  incomes  and  lesser  deprivation  are  considered  higher  socioeconomic  status  

   

 

36

Table 2.3. SES-related Relative Concentration Indices in the use of eye care services, 1995-2009

  RCI(income)   Change   RCI(MDI  )   Change   RCI(SDI)   Change  

1995   0.0309     0.0889     -­‐0.0657    1996   0.0364   0.005   0.0872   -­‐0.002   -­‐0.0557   0.010  

1997   0.0272   -­‐0.009   0.0766   -­‐0.011   -­‐0.0512   0.005  

1998   0.0326   0.005   0.0772   0.001   -­‐0.0484   0.003  

1999   0.0310   -­‐0.002   0.0720   -­‐0.005   -­‐0.0415   0.007  

2000   0.0272   -­‐0.004   0.0680   -­‐0.004   -­‐0.0477   -­‐0.006  

2001   0.0282   0.001   0.0627   -­‐0.005   -­‐0.0349   0.013  

2002   0.0273   -­‐0.001   0.0606   -­‐0.002   -­‐0.0360   -­‐0.001  

2003   0.0262   -­‐0.001   0.0588   -­‐0.002   -­‐0.0347   0.001  

2004   0.0261   0.000   0.0535   -­‐0.005   -­‐0.0319   0.003  

2005   0.0229   -­‐0.003   0.0530   -­‐0.001   -­‐0.0362   -­‐0.004  

2006   0.0245   0.002   0.0492   -­‐0.004   -­‐0.0322   0.004  

2007   0.0300   0.006   0.0529   0.004   -­‐0.0308   0.001  

2008   0.0270   -­‐0.003   0.0482   -­‐0.005   -­‐0.0253   0.006  

2009   0.0214   -­‐0.006   0.0427   -­‐0.005   -­‐0.0278   -­‐0.003  

 

 

   

 

37

Tab

le 2

.4. G

eogr

aphi

c de

com

posi

tion

of S

ES-r

elat

ed C

once

ntra

tion

Indi

ces

by A

HS

zone

s

!

!!!!!!!!!!!!!!!!!!!!!!!!!!!

Income'

Material'deprivation'

Social'deprivation'

RCI(B)'

RCI(W)'

Residual'

Total'RCI'

RCI(B)'RCI(W)'

Residual'Total'RCI'

RCI(B)'

RCI(W)'

Residual'Total'RCI'

1995'

0.0156!

0.0053!

0.0100!

0.0309%

0.0802!

0.0130!

*0.0043!

0.0889%

*0.0578!

*0.0093!

0.0014!

'0.0657%

1996'

0.0261!

0.0059!

0.0043!

0.0364%

0.0841!

0.0176!

*0.0145!

0.0872%

*0.0567!

*0.0024!

0.0034!

'0.0557%

1997'

0.0149!

0.0057!

0.0066!

0.0272%

0.0750!

0.0168!

*0.0152!

0.0766%

*0.0547!

*0.0018!

0.0053!

'0.0512%

1998'

0.0108!

0.0070!

0.0149!

0.0326%

0.0686!

0.0173!

*0.0087!

0.0772%

*0.0561!

*0.0015!

0.0091!

'0.0484%

1999'

0.0086!

0.0065!

0.0159!

0.0310%

0.0627!

0.0163!

*0.0069!

0.0720%

*0.0482!

*0.0009!

0.0076!

'0.0415%

2000'

0.0008!

0.0059!

0.0205!

0.0272%

0.0561!

0.0164!

*0.0044!

0.0680%

*0.0420!

*0.0025!

*0.0032!

'0.0477%

2001'

0.0095!

0.0072!

0.0116!

0.0282%

0.0580!

0.0156!

*0.0109!

0.0627%

*0.0412!

0.0005!

0.0058!

'0.0349%

2002'

0.0184!

0.0050!

0.0039!

0.0273%

0.0623!

0.0141!

*0.0158!

0.0606%

*0.0407!

*0.0005!

0.0052!

'0.0360%

2003'

0.0290!

0.0041!

*0.0069!

0.0262%

0.0670!

0.0135!

*0.0216!

0.0588%

*0.0359!

*0.0008!

0.0019!

'0.0347%

2004'

0.0242!

0.0043!

*0.0025!

0.0261%

0.0562!

0.0134!

*0.0161!

0.0535%

*0.0235!

*0.0011!

*0.0073!

'0.0319%

2005'

0.0223!

0.0039!

*0.0033!

0.0229%

0.0526!

0.0133!

*0.0130!

0.0530%

*0.0244!

*0.0022!

*0.0096!

'0.0362%

2006'

0.0216!

0.0039!

*0.0010!

0.0245%

0.0446!

0.0131!

*0.0085!

0.0492%

*0.0139!

*0.0017!

*0.0167!

'0.0322%

2007'

0.0249!

0.0049!

0.0002!

0.0300%

0.0419!

0.0142!

*0.0032!

0.0529%

*0.0099!

*0.0018!

*0.0190!

'0.0308%

2008'

0.0250!

0.0040!

*0.0020!

0.0270%

0.0383!

0.0139!

*0.0040!

0.0482%

*0.0034!

*0.0012!

*0.0206!

'0.0253%

2009'

0.0216!

0.0027!

*0.0029!

0.0214%

0.0328!

0.0083!

0.0017!

0.0427%

*0.0021!

*0.0063!

*0.0194!

'0.0278%

38

Figure 2.1. Relative Concentration Index in eye care services by ophthalmologists

Figure 2.2. The trends of household income-related RCIs in eye care services by ophthalmologists

-0.0800

-0.0600

-0.0400

-0.0200

0.0000

0.0200

0.0400

0.0600

0.0800

0.1000

199519961997199819992000200120022003200420052006200720082009

Household Income CI Material Deprivation CI

Social Deprivation CI

-0.0200

0.0000

0.0200

0.0400

0.0600

0.0800

0.1000

0.1200

0.1400

199519961997199819992000200120022003200420052006200720082009

South Calgary Central Edmonton North

 

39

Figure 2.3. The trends of material deprivation index-related RCIs in eye care services by ophthalmologists

 

 

Figure 2.4. The trends of social deprivation index-related RCIs in eye care services by ophthalmologists

 

 

-­‐0.0200    

0.0000    

0.0200    

0.0400    

0.0600    

0.0800    

0.1000    

1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  South   Calgary   Central   Edmonton   North  

-­‐0.1000    

-­‐0.0800    

-­‐0.0600    

-­‐0.0400    

-­‐0.0200    

0.0000    

0.0200    

0.0400    

1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  

South   Calgary   Central   Edmonton   North  

40

References

1. Sanmartin C, Gilmore J. Diabetes care in Canada: Results from selected provinces

and territory, 2005. Ottawa: Statistics Canada, 2006.

2. Buhrmann R, Houdege W, Beardmore J, Baker G, Lowcock B, Pan I, & Bovell A.

Chapter 2. The Epidemilogoy of Blindness and Visual Imparment in Canada :

Diabetic Retinopathy. Foundations for a Canadian Vision Health Strategy : Towards

Preventing Avoidable Blindness and Promoting Vision Health: The National

Coalition for Vision Health; 2007. p. 14-5.

3. Canadian Diabetes Association. Rethinopahty: Canadian Diabetes Association 2008

clinical practice guidelines for the prevention and management of diabetes in Canada.

Can J Diabetes. 2008;32(suppl2):S134-S9.

4. Jones S, Edwards RT. Diabetic retinopathy screening: a systematic review of the

economic evidence. Diabet Med. 2010;27(3):249-56.

5. Kobrin Klein BE. Overview of Epidemiologic Studies of Diabetic Retinopathy.

Ophthalmic epidemiolo. 2007;14(4):179-83.

6. American Academy of Ophthalmology. Chapter 5 : Retinal Vascular Disease. Retina

and Vitreous Section12 Basic and Clinical Science Course 2011-2012. San Fransico,

CA: American Academy of Ophthalmology The Eye M.D. Association; 2011. p. 107-

80.

7. Public Health Task Force for Vision and Ophthalmology. Determining the

information needs for prevention of vision loss strategies in Canada. Public Health

Task Force for Vision and Ophthalmology. Can J Ophthalmol. 2000;35(5):255-7.

8. Szabo SM, Beusterien KM, Pleil AM, Wirostko B, Potter MJ, Tildesley H, et al.

Patient Preferences for Diabetic Retinopathy Health States. Invest Ophthalmol Vis

Sci. 2010;51(7):3387-94.

9. Aiello LP, Cavallerano J, & Bursell SE. Diabetic eye disease. Endocrinol Metab Clin

North Am. 1996;25(2):272-90.

10. Perruccio AV, Badley EM, & Trope GE. A Canadian population-based study of

vision problems: assessing the significance of socioeconomic status. Can J

Ophthalmol. 2010;45(5):477-83.

11. Hooper P, Boucher MC, Cruess A, Dawson KG, Delpero W, & Greve M, et al.

Canadian Ophthalmological Society Evidence-Based Clinical Practice Guidelines for

the Management of Diabetic Retinopathy – Executive Summary. Can J Ophthalmol.

2012;47(2):91-101.

12. van Eijk KN, Blom JW, Gussekloo J, Polak BC, & Groeneveld Y. Diabetic

retinopathy screening in patients with diabetes mellitus in primary care: Incentives

and barriers to screening attendance. Diabetes Res Clin Pract. 2012;96(1):10-6.

41

13. Rudnisky CJ, Tennat MTS, Johnson JA, & Balko SU. Chapter 9 : Diabetes and Eye

Disease in Alberta. In: Johnson JA,editor. Alberta Diabetes Atlas 2011. Edmonton:

Institute of Health Economics 2011. p. 153-74.

14. Zhang X, Andersen R, Saaddine JB, Beckles GL, Duenas MR, & Lee PP. Measuring

access to eye care: a public health perspective. Ophthalmic epidemiol.

2008;15(6):418-25.

15. Carmen D, Harris LP, & Mukhtar Q. Changes in receiving preventive care services

among US adults with diabetes, 1997-2007. Prev Chronic Dis. 2010;7(3):A56.

16. Moss SE, Klein R, Klein BEK. Factors Associated With Having Eye Examinations in

Persons With Diabetes. Arch Fam Med. 1995;4(6):529-34.

17. Burhrmann R, Assaad D, Hux JE, Tang M, & Sykora K. Chapter 10: Diabetes and

the Eye. Diabetes in Ontario: An ICES Practice Atlas: Institute for Clinical Evaluative

Sciences; 2003. p. 10.193- 10.208.

18. Kliner M, Fell G, Gibbons C, Dhothar M, Mookhtiar M, & Cassels-Brown A.

Diabetic retinopathy equity profile in a multi-ethnic, deprived population in Northern

England. Eye (Lond). 2012.

19. Scanlon PH, Carter SC, Foy C, Husband RF, Abbas J,& Bachmann MO. Diabetic

retinopathy and socioeconomic deprivation in Gloucestershire. J Med Screen.

2008;15(3):118-21.

20. Millett C, Dodhia H. Diabetes retinopathy screening: audit of equity in participation

and selected outcomes in South East London. J Med Screen. 2006;13(3):152-5.

21. Dunlop S, Coyte PC, & McIsaac W. Socio-economic status and the utilisation of

physicians' services: results from the Canadian National Population Health Survey.

Social Sci & Med. 2000;51(1):123-33.

22. Goyder EC, McNally PG, &Botha JL. Inequalities in access to diabetes care:

evidence from a historical cohort study. Qual Health Care. 2000;9(2):85-9.

23. Bowen S. Part 1. Equity in Access to Health Care : Document 1. Access to Health

Services for Underserved Population in Canada. Certain Circumstances : Issues in

Equity and Responsiveness in Access to Health Care in Canada: Health Canada; 2001.

p. 7-60.

24. The Nuffield Trust. Equity in Health and Healthcare. In Oliver A, ed: The Nuffield

Trust for Research and Policy Studeise in Health Services; 2003.

25. Organization for Economic Co-operation and Development. Growing unequal:

income distribution and poverty in OECD Nations. Paris: Organization for Economic

Co-operation and Development. 2008.

26. Allin S. Does Equity in Healthcare Use Vary across Canadian Provinces? Healthcare

Policy. 2008;3(4):83-99.

42

27. Alberta Health and Wellness. Alberta Diabetes Strategy, 2003-2013: Alberta Health

and Wellness; 2003.

28. Johnson JA, Balko SU, Hugel G. Chapter 1 : Background and methods. In: Johnson

JA, ed. Alberta Diabetes Atlas 2011. Edmonton: Institute of Health Economics; 2011.

p. 3-12.

29. Pampalon R, Hamel D, Gamache P, & Raymond G. A deprivation index for health

planning in Canada. Chronic diseases in Canada. 2009;29(4):178-91.

30. Kakwani N. Income Inequality and Poverty : Methods of Estimation and Policy

Applications. New York: Oxford University Press; 1980.

31. Wagstaff A, Paci P, & van Doorslaer E. On the measurement of inequalities in health.

Social Sci Med. 1991;33(5):545-57.

32. O'Donnell O, van Dooslaer E, Wagstaff A, & Lindelow M. Chapter 7. Concentration

Curves. Analyzing Health Equity Using Household Survey Data : A Guide to

Techniques and Their Implementation. Washington, DC: The World Bank; 2009. p.

83-94.

33. O'Donnell O, van Dooslaer E, Wagstaff A, & Lindelow M. Chapter 8. The

Concentration Index. Analyzing Health Equity Using Household Survey Data : A

Guide to Techniques and Their Impementation. Washington, DC: The World Bank;

2008. p. 95-108.

34. Wagstaff A. The bounds of the concentration index when the variable of interest is

binary, with an application to immunization inequality. Health economics.

2005;14(4):429-32.

35. Haper S, Lynch J. Methods for Measuring Cancer Disparities : Using Data Relevant

to Healthy people 2010 Cancer-Related Objectives. Bethesda, MD: National Cancer

Institute; 2005.

36. Johnson JA, Balko SU, Hugel G, Low C, Svenson LW. Increasing incidence and

prevalence with limited survival gains among rural Albertans with diabetes: a

retrospective cohort study, 1995–2006. Diabet Med. 2009;26(10):989-95.

37. Clarke PM, Gerdtham UG, & Connelly LB. A note on the decomposition of the

health concentration index. Health economics. 2003;12(6):511-6.

38. Wagstaff A. Inequality decomposition and geographic targeting with applications to

China and Vietnam. Health economics. 2005;14(6):649-53.

39. Morris S, Sutton M, & Gravelle H. Inequity and inequality in the use of health care in

England: an empirical investigation. Social Sci & Med. 2005;60(6):1251-66.

40. Lorant V, Boland B, Humblet P, & Deliege D. Equity in prevention and health care. J

Epidemiol Community Health. 2002;56(7):510-6.

41. Lemstra M, Neudorf C, & Opondo J. Health disparity by neighbourhood income. Can

43

J Public Health. 2006;97(6):435-9.

42. van Doorslaer E, Masseria C. Income-Related Inequality in the Use of Medical Care

in 21 OECD Countries. Paris: Organization for Economic Co-operation and

Development. 2004.

43. Veugelers PJ, Yip AM. Socioeconomic disparities in health care use: Does universal

coverage reduce inequalities in health? J Epidemiol Community Health.

2003;57(6):424-8.

44. Nam S, Chesla C, Stotts NA, Kroon L, & Janson SL. Barriers to diabetes

management: Patient and provider factors. Diabetes Res Clin Pract 2011;93(1):1-9.

45. Jotkowitz AB, Rabinowitz G, Segal AR, Weitzman R, Epstein L, Porath A. Do

Patients with Diabetes and Low Socioeconomic Status Receive Less Care and Have

Worse Outcomes? A National Study. Am J Med 2006;119(8):665-9.

46. Perry M, Williams RL, Wallerstein N, & Waitzkin H. Social Capital and Health Care

Experiences Among Low-Income Individuals. Am JPublic Health. 2008;98(2):330-6.

47. Wilson K, Rosenberg MW. The geographies of crisis: exploring accessibility to

health care in Canada. Can Geograph. 2002;46(3):223-34.

48. Sibley LM, Weiner JP. An evaluation of access to health care services along the rural-

urban continuum in Canada. BMC Health Serv Res. 2011;11:20.

49. College of Physician and Surgeons of Alberta. College of Physicians and Surgeons of

Alberta - Ophthalmologist (Internet). [cited 2013 March, 13th]; Available from:

http://www.cpsa.ab.ca/PhysicianSearch/SearchResults.aspx?Specialty=Ophthalmology.

50. Alberta Health Services. Demographics. Alberta Health Services; [cited 2013 March,

17th]; Available from: http://www.albertahealthservices.ca/745.asp.

51. van Doorslaer E, Jones AM. Inequalities in self-reported health: validation of a new

approach to measurement. J Health Econ. 2003;22(1):61-87.

52. Odoi A, Wray R, Emo M, Birch S, Hutchison B, Eyles J, et al. Inequalities in

neighbourhood socioeconomic characteristics: potential evidence-base for

neighbourhood health planning. International J Health Geograph. 2005;4(1):20.

 

44

Supplementary Tables and Figures

Table S2.1. Alberta physician claims data

Procedure   Code   Description  

Retinal  Laser  Treatment  (Retinal  Photocoagulation)  

28.5A   Focal  and/or  pan  retinal  photocoagulation  

Vitrectomy  Surgery   28.72A   Vitreous  cavity  washout  

  28.72B   Total  vitrectomy  

  28.74A   Discission  of  vitreous/retinal  adhesions  

  28.74B  

Stripping  of  premacular  membrane,  Associated  vitrectomy  and  retinal  encircling  

Cataract  Surgery   27.72  Insertion  of  intraocular  lens  prosthesis  with  cataract  extraction,  one-­‐stage  

 Laser  Treatment  for  Glaucoma  

26.52  (1995-­‐2004)   Laser  peripheral  iridotomy  

26.52A  Either  laser  peripheral  iridotomy  or  argon  laser  trabeculoplasty  or  selective  laser  trabeculoplasty  

End-­‐Stage  Glaucoma  (Laser  Treatment)   26.98B(2005-­‐2009)  

Diode  laser  cyclophotocoagulation  (ciliary  body  ablation)  

 Glaucoma  Surgery   26.2A  (1995-­‐2009)   Major  glaucoma  operation  

(trabeculectomy,  EMS  shunt)  

26.2B  (1995-­‐2009)   Ahmed  shunt  or  Baerveldt  shunt,  with  scleral  patch  graft  

26.25A  (1998-­‐2009)   Repeat  trabeculectomy  

 

Source:  Alberta  Diabetes  Atlas  2011  

 

 

 

 

 

 

 

45

 

Tab

le S

2.2.

Nor

th z

one

– D

istri

butio

n of

eye

car

e se

rvic

es b

y op

htha

lmol

ogis

ts fo

r Abo

rigin

al a

nd n

on-A

borig

inal

peo

ple

with

dia

bete

s, 1

995-

2009

!

Eye$care$service$

1995$

1996$1997$1998$1999$2000$2001$2002$2003$2004$2005$2006$2007$2008$2009$

Aboriginal$

Yes!

144!

167!

204!

217!

252!

272!

363!

502!

552!

684!

687!

702!

685!

653!

732!

No!

942!

1,009!1,063!1,137!1,243!1,370!1,435!1,465!1,602!1,591!1,739!1,870!2,050!2,220!2,287!

%!13.3!

14.2!

16.1!

16.0!

16.9!

16.6!

20.2!

25.5!

25.6!

30.1!

28.3!

27.3!

25.0!

22.7!

24.2!

Non<

Aboriginal$

Yes!

1,309!

1,559!1,725!1,943!2,211!2,315!2,676!2,865!3,208!3,496!3,718!4,042!4,353!4,698!5,100!

No!

7,142!

7,281!7,494!7,778!8,207!8,809!9,335!1

0,135!10,859!11,499!12,269!12,888!13,565!14,195!15,148!

%!15.5!

17.6!

18.7!

20.0!

21.2!

20.8!

22.3!

22.0!

22.8!

23.3!

23.3!

23.9!

24.3!

24.9!

25.2!

 

46

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Tab

le S

2.3.

Nor

th z

one

– D

istri

butio

n of

eye

car

e se

rvic

es in

Abo

rigin

al p

eopl

e w

ith d

iabe

tes

by in

com

e qu

intil

e

Income'

Quintile'

1995'

1996'

1997'

1998'

1999'

2000'

2001'

2002'

2003'

2004'

2005'

2006'

2007'

2008'

2009'

1st '(Lower)'

12.0%

12.9%

15.5%

12.5%

13.8%

16.9%

19.0%

27.9%

26.5%

31.1%

30.2%

33.9%

31.7%

28.7%

28.8%

2nd ''

14.8%

11.9%

15.5%

18.9%

16.4%

14.2%

18.3%

18.4%

19.3%

26.7%

19.8%

20.5%

18.0%

15.4%

19.9%

3rd ''

14.4%

15.0%

16.5%

18.3%

17.4%

16.6%

18.4%

23.7%

28.3%

31.4%

32.7%

27.0%

22.7%

19.8%

19.3%

4th ''

18.0%

21.7%

21.5%

19.2%

24.2%

25.3%

24.8%

28.5%

24.9%

25.9%

27.8%

25.3%

22.6%

22.9%

25.4%

5th '(Higher)'

9.4%

12.7%

14.0%

13.4%

16.8%

14.1%

24.1%

30.6%

27.8%

33.7%

28.7%

27.0%

29.4%

26.7%

29.0%

RCI'

C0.024'

0.047'

0.014'

0.015'

0.057'

0.003'

0.046'

0.029'

0.017'

0.014'

0.011'

C0.042'C0.013'C0.007'

0.010'

%

Income'

Quintile'

1995'

1996'

1997'

1998'

1999'

2000'

2001'

2002'

2003'

2004'

2005'

2006'

2007'

2008'

2009'

1st '(Lowe

r)'

15.0%

17.8%

18.5%

18.3%

19.9%

19.0%

19.8%

20.0%

20.4%

21.3%

21.3%

20.6%

20.5%

20.2%

19.2%

2nd ''

16.0%

17.0%

18.6%

20.9%

20.9%

19.4%

19.4%

16.8%

16.7%

18.2%

17.7%

16.7%

16.5%

15.2%

14.4%

3rd ''

15.7%

19.3%

19.1%

20.8%

24.1%

23.0%

28.0%

29.3%

31.5%

31.0%

31.1%

29.9%

29.6%

29.8%

28.2%

4th''

16.6%

19.0%

20.5%

22.1%

22.9%

24.6%

27.6%

28.5%

30.8%

31.2%

31.3%

33.2%

33.4%

35.5%

42.1%

5th'(high

er)'

10.7%

12.4%

12.4%

13.3%

16.4%

14.9%

12.6%

13.2%

11.4%

10.5%

10.7%

16.0%

20.5%

23.1%

23.5%

RCI'

B0.00

7'B0.00

5'0.0

06'

0.010'

0.019'

0.035'

0.042'

0.064'

0.071'

0.055'

0.060'

0.091'

0.109'

0.135'

0.130'

%Tab

le S

2.4.

Nor

th z

one-

Dis

tribu

tion

of e

ye c

are

serv

ices

in n

on-A

borig

inal

s by

inco

me

quin

tile  

 

47

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  MDI$Quintile$

1995$

1996$

1997$

1998$

1999$

2000$

2001$

2002$

2003$

2004$

2005$

2006$

2007$

2008$

2009$

1st $(More)$

11.5$

12.1$

14.4$

11.8$

14.8$

16.4$

21.7$

30.7$

28.5$

34.1$

32.2$

33.2$

33.0$

29.5$

31.1$

2nd $$

14.8$

11.9$

15.5$

18.9$

16.4$

14.2$

18.3$

18.4$

19.3$

26.7$

19.8$

20.5$

18.0$

15.4$

19.9$

3rd $$

18.2$

12.8$

12.9$

18.9$

20.4$

17.1$

19.1$

26.1$

23.5$

25.8$

25.6$

29.8$

21.6$

20.0$

25.2$

4th $$

14.9$

19.4$

16.5$

35.3$

17.4$

24.4$

18.4$

23.7$

28.2$

31.3$

32.7$

27.0$

22.6$

20.6$

19.3$

5th $(Less)$

11.0$

19.4$

20.5$

11.7$

19.5$

17.5$

23.1$

25.5$

23.4$

25.1$

19.1$

21.8$

23.0$

22.2$

24.4$

RCI$

0.002$

0.088$

0.073$

0.064$

0.049$

0.026$

A0.023$A0.055$A0.028$A0.044$A0.029$A0.069$A0.079$A0.075$A0.079$

!

MDI$Quintile$

1995$

1996$

1997$

1998$

1999$

2000$

2001$

2002$

2003$

2004$

2005$

2006$

2007$

2008$

2009$

1st $(More)$

13.6%

15.9%

16.3%

18.3%

18.9%

19.4%

23.3%

22.6%

22.9%

24.4%

24.0%

24.1%

24.1%

23.5%

24.6%

2nd $$

16.2%

19.7%

20.0%

21.0%

22.0%

19.7%

18.9%

17.2%

17.7%

18.7%

18.3%

17.0%

16.5%

15.5%

14.9%

3rd $$

15.9%

17.6%

18.3%

21.0%

23.4%

22.0%

25.0%

26.3%

28.4%

27.7%

26.6%

26.7%

26.8%

28.0%

26.9%

4th $$

17.4%

19.5%

21.7%

22.3%

23.2%

25.0%

28.0%

29.2%

31.1%

31.4%

31.8%

33.2%

33.4%

35.5%

42.1%

5th $(Less)$

10.9%

12.8%

11.8%

12.0%

16.0%

14.6%

11.6%

12.5%

10.8%

9.4%

10.5%

15.6%

20.5%

22.9%

23.4%

RCI$

0.023$

0.018$

0.026$

0.014$

0.028$

0.035$

0.018$

0.042$

0.049$

0.025$

0.039$

0.062$

0.083$

0.109$

0.101$

!

Tab

le S

2.5.

Nor

th z

one-

Dis

tribu

tion

of e

ye c

are

serv

ices

in A

borig

inal

s by

MD

I qui

ntile

Tab

le S

2.6.

Nor

th z

one-

Dis

tribu

tion

of e

ye c

are

serv

ices

in n

on- A

borig

inal

s by

MD

I qui

ntile  

 

48

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Eye$care$Service$

1995$1996$1997$1998$1999$2000$2001$2002$2003$2004$2005$2006$2007$2008$2009$

Urban$

Yes$

436$

543$

731$

812$

973$

1,036$1

,202$

1,289$1,468$1

,567$

1,695$1

,917$

2,184$2

,421$

2,660$

No$2,333$2

,430$

2,977$3

,255$

3,516$3

,878$

4,217$4

,673$

5,085$5

,531$

5,982$6

,266$

6,647$6

,956$

7,687$

%$15.7$

18.3$

19.7$

20.0$

21.7$

21.1$

22.2$

21.6$

22.4$

22.1$

22.1$

23.4$

24.7$

25.8$

25.7$

Rural$

Yes$1,017$1

,183$

1,198$1

,348$

1,490$1

,551$

1,837$2

,078$

2,292$2

,613$

2,710$2

,827$

2,854$2

,930$

3,172$

No$5,751$5

,860$

5,580$5

,660$

5,934$6

,301$

6,553$6

,927$

7,376$7

,559$

8,026$8

,492$

8,968$9

,459$

9,748$

%$15.0$

16.8$

17.7$

19.2$

20.1$

19.8$

21.9$

23.1$

23.7$

25.7$

25.2$

25.0$

24.1$

23.7$

24.6$

!Tab

le S

2.7.

Nor

th z

one

– D

istri

butio

n of

eye

car

e se

rvic

es b

y op

htha

lmol

ogis

ts fo

r urb

an a

nd ru

ral d

wel

ling

peop

le w

ith d

iabe

tes,

1995

-200

9  

 

49

Tab

le S

2.8.

Nor

th z

one-

Dis

tribu

tion

of e

ye c

are

serv

ices

in u

rban

dw

elle

rs b

y in

com

e qu

intil

e  

Tab

le S

2.9.

Nor

th z

one-

Dis

tribu

tion

of e

ye c

are

serv

ices

in ru

ral d

wel

lers

by

inco

me

quin

tile  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Income'

Quintile'

1995'

1996'1997'

1998'1999'

2000'2001'

2002'2003'

2004'

2005'

2006'2007'2008'

2009'

1st '(Lower)'

14.7%

19.2%

20.5%

20.2%

20.6%

18.7%

19.7%

17.9%

18.2%

17.5%

16.9%

17.2%

16.8%

16.0%

15.0%

2nd ''

16.0%

21.4%

21.7%

23.4%

25.2%

22.5%

23.7%

20.3%

21.1%

20.8%

20.6%

18.3%

18.6%

16.1%

13.1%

3rd ''

15.6%

15.3%

17.0%

17.1%

16.5%

18.8%

18.5%

16.4%

14.8%

16.6%

15.6%

21.4%

20.0%

23.4%

24.7%

4th ''

14.8%

24.8%

27.6%

27.4%

30.6%

30.3%

36.3%

39.4%

44.5%

43.3%

43.8%

41.0%

42.4%

44.4%

43.1%

5th '(Higher)'

15.4%

12.7%

11.9%

11.8%

15.5%

14.3%

11.8%

12.6%

11.3%

10.0%

10.5%

15.4%

20.8%

22.5%

23.3%

RCI'

C0.028'

C0.034'

C0.052'

C0.050'

C0.011'

0.003'

0.000'

0.032'

0.032'

0.018'

0.031'

0.063'

0.095'

0.120'

0.118'

!

Income'

Quintile'

1995'

1996'

1997'

1998'

1999'

2000'

2001'

2002'

2003'

2004'

2005'

2006'

2007'

2008'

2009'

1st '(Lower)'

14.5%

17.0%

17.1%

16.9%

18.3%

18.8%

19.9%

22.3%

22.4%

24.6%

24.8%

24.8%

24.1%

23.4%

23.3%

2nd ''

16.1%

16.0%

16.4%

18.4%

18.1%

18.0%

20.4%

18.9%

18.5%

20.1%

18.7%

17.1%

17.8%

16.5%

19.0%

3rd ''

15.6%

16.1%

17.6%

20.1%

20.0%

18.0%

17.8%

17.5%

18.8%

22.9%

21.8%

20.6%

18.2%

16.8%

17.1%

4th ''

15.7%

19.1%

19.2%

20.6%

24.1%

23.0%

26.8%

29.1%

31.1%

30.3%

30.7%

31.0%

30.5%

31.1%

32.6%

5th '(Higher)'

13.8%

16.5%

18.5%

21.2%

20.8%

21.9%

26.8%

29.3%

29.6%

31.5%

30.7%

31.5%

30.7%

31.1%

32.1%

RCI'

C0.010'

C0.009'

0.018'

0.038'

0.038'

0.040'

0.069'

0.079'

0.085'

0.069'

0.073'

0.086'

0.083'

0.095'

0.102'

!

 

50

Tab

le S

2.10

. Nor

th z

one-

Dis

tribu

tion

of e

ye c

are

serv

ices

in u

rban

dw

elle

rs b

y M

DI q

uint

ile  

Tab

le S

2.11

. Nor

th z

one-

Dis

tribu

tion

of e

ye c

are

serv

ices

in ru

ral d

wel

lers

by

MD

I qui

ntile  

 

 

 

 

 

 

 

 

 

 

MDI$Quintile$

1995$

1996$

1997$

1998$

1999$

2000$

2001$

2002$

2003$

2004$

2005$

2006$

2007$

2008$

2009$

1st $(Mo

re)$

13.3$

18.8$

19.4$

20.2$

21.0$

19.7$

22.5$

20.5$

21.0$

20.2$

19.9$

19.6$

18.9$

17.5$

17.4$

2nd $$

16.3$

20.8$

24.8$

24.6$

26.1$

21.4$

19.5$

16.2$

16.6$

16.2$

16.3$

14.3$

14.8$

13.7$

11.6$

3rd $$

15.7$

15.7$

17.1$

17.3$

16.8$

19.3$

18.8$

16.5$

15.2$

17.2$

15.7$

21.4$

20.4$

23.4$

24.6$

4th$$

15.7$

24.8$

27.6$

27.4$

30.6$

30.3$

36.3$

39.4$

44.5$

43.3$

43.8$

41.0$

42.4$

44.4$

43.1$

5th$(Less)$

15.6$

12.7$

11.9$

11.8$

15.4$

14.3$

11.8$

12.6$

11.2$

10.0$

10.5$

15.4$

20.8$

22.6$

23.2$

RCI$

@0.02

3$@0.03

4$@0.03

6$@0.05

1$@0.00

7$@0.00

3$@0.01

8$0.0

17$

0.016$

0.006$

0.016$

0.050$

0.085$

0.114$

0.108$

!

MDI$Quintile$

1995$

1996$

1997$

1998$

1999$

2000$

2001$

2002$

2003$

2004$

2005$

2006$

2007$

2008$2009$

1st $(More)$

13.3$

14.5$

15.7$

16.1$

17.1$

17.3$

20.8$

22.5$

21.7$

25.3$

23.4$

23.8$

23.7$

22.7$

24.4$

2nd $$

16.6$

19.2$

18.6$

21.0$

22.1$

21.7$

22.7$

22.4$

24.4$

25.1$

25.8$

24.5$

23.5$

23.2$

22.1$

3rd $$

15.8$

16.0$

17.8$

20.0$

19.8$

17.9$

17.3$

17.8$

19.1$

22.2$

22.1$

20.3$

17.7$

16.3$

16.6$

4th$$

15.7$

18.3$

18.4$

21.2$

23.4$

21.9$

25.0$

27.0$

28.2$

27.7$

26.9$

26.6$

26.2$

27.4$

29.1$

5th$(Less)$

14.4$

18.9$

20.7$

21.4$

21.7$

24.6$

27.9$

30.1$

31.0$

31.0$

32.9$

34.5$

33.9$

33.9$

35.2$

RCI$

0.039$0.0

46$0.0

48$0.0

60$0.0

56$0.0

60$0.0

38$0.0

32$0.0

49$0.0

07$0.0

38$0.0

34$0.0

28$0.0

40$0.0

40$

!

 

51

Figure S2.1. Proportion of Aboriginals and non- Aboriginals in North Zone with eye care services by ophthalmologists

 

 

Figure S2.2. Income-related RCI by Aboriginal status in North Zone

 

 

 

 

 

0%  

5%  

10%  

15%  

20%  

25%  

30%  

35%  

1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  

Aboriginal   Non-­‐Aboriginal  

-­‐0.100  

-­‐0.050  

0.000  

0.050  

0.100  

0.150  

1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  

Aboriginal   Non-­‐Aboriginal  

 

52

Figure S2.3. MDI-related RCI by Aboriginal status in North Zone

 

 

Figure S2.4.  Proportion of urban and rural dwellers in North Zone with eye care services by an ophthalmologist  

 

-­‐0.100  

-­‐0.050  

0.000  

0.050  

0.100  

0.150  

1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  

Aboriginal   Non-­‐Aboriginal  

0%  

5%  

10%  

15%  

20%  

25%  

30%  

35%  

1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  

Urban   Rural  

 

53

Figure S2.5. Income-related RCI by Urban and rural in North zone

 

 

 

Figure S2.6. MDI-related RCI by urban and rural in North Zone

 

-­‐0.100  

-­‐0.050  

0.000  

0.050  

0.100  

0.150  

1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  

Urban   Rural  

-­‐0.100  

-­‐0.050  

0.000  

0.050  

0.100  

0.150  

1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  

Urban   Rural  

   

54  

Chapter 3. Determinants of Socioeconomic Inequities in the Use of Eye Care Services among Diabetic Patients in Alberta, Canada 1995-2009

Abstract

Diabetic retinopathy (DR) is a serious life-threatening complication in diabetic patients.

The Canadian Diabetes Association clinical guidelines recommend an annual dilated eye

examination by an eye care specialist for timely detection and effective prevention.

Previous studies have suggested that socioeconomic status (SES) is associated with use of

the recommended eye care services. However, Canadian evidence on the factors

associated with SES-related disparities in the use of eye screening services among

diabetic patients is limited. Therefore, the aim of this study was to assess SES-related

disparities in the use of eye care services among diabetic patients in Alberta, Canada. We

applied the econometric techniques of Relative Concentration Index (RCI), and used

regression-based decomposition analysis to identify major contributors to SES-related

disparities. Horizontal inequity index (HI), which represents equal access for equal need,

was calculated based on decomposition methods by quantifying contributions of need and

non-need factors. SES was represented by 3 different measures: census-based median

household income, and material (MDI) and social deprivation (SDI) indices. This study

used data from the Alberta Diabetes Surveillance System (ADSS) 1995-2009: a total of

1,949,498 diabetic patients over a 15-year period were included in the analyses. Eye care

service was defined in this study as any visit to an ophthalmologist, based on the medical

services claims database. We found horizontal inequity among diabetic patients in the use

of eye care services by an ophthalmologist but these differed depending on the specific

SES indicator. Income and material deprivation-related HIs have been in favour of richer

groups (i.e., pro-rich), however, the social deprivation-related HIs have been in favour of

poorer groups (i.e. socially deprived diabetic patients tended to use more eye screening

   

55  

services than those who were less socially deprived). In addition, the study found that the

MDI and place of residence (urban/ rural) were important contributors to the observed

“pro-rich” income- and MDI-related RCIs. The observed SDI-related inequity was

explained by SDI itself. The findings imply that economic- and social-related resources

generate different directions of disparities in the use of eye screening services and also

suggest the need for developing health policy to alleviate different determinants of SES-

related disparities in the use of eye screening services in diabetic patients in Alberta.

   

56  

3.1. Introduction

Diabetic retinopathy (DR) is a major cause of visual impairment that ultimately impedes

daily activities and decreases quality of life (1-5). The development of DR is closely

associated with the duration of diabetes: as the diabetic condition develops, the risk of

visual impairment increases (1). Approximately 78% of patients with a 15-year duration

of diabetes have DR, which indicates higher prevalence of DR compared to diabetic

patients with 5-year duration (6). Visual impairment caused by DR is preventable and

manageable if diagnosed at the early stage of its course (1, 7). Thus, diabetic patients with

sight-threatening DR should receive eye examinations by either an ophthalmologist or a

retinal specialist as part of diabetic eye complications management (7). For early

detection and timely treatment, it is recommended that all diagnosed diabetic patients

receive periodic eye care services by a health professional (7).

Despite well-documented and comprehensive clinical guidelines, there is underuse of the

recommended eye care services among diabetic patients in Canada. One study of five

provinces and one territory reported that only 68% of patients with diabetes had received

the recommended level of eye examination services (8). In Alberta, only 49% of patients

received an eye examination by an ophthalmologist within three years of being diagnosed

with diabetes and the rate of eye examinations has decreased over previous years (3).

Equal access to health care is enshrined in the Canada Health Act (CHA), clearly

indicating that equity in health care is a main goal of the legislation and that no one

should be discriminated against on the basis of their age, sex, income, or educational

attainment (9, 10). The CHA states that medically necessary services for maintaining

health, preventing disease or diagnosing or treating an illness should be provided by

   

57  

provincial health care plan to all residents without barriers, eye care services for diabetic

patients fall under medically necessary services (7, 10). Nonetheless, a growing body of

literature has identified several barriers to eye care service access in the general Canadian

population (11, 12). These include limited resources to meet the demand for eye care

services, variations in health care coverage between provinces, and lack of collaborative

services between health care providers (11). While differences in health care utilization

amongst patients may be driven by clinical need, differences that are driven by non-need

factors, such as socioeconomic status (SES) or geography represent inequities, are not

simply disparities. The presence of disparities that are drive by non-need factors is

referred to as horizontal inequity index (HI), the amelioration of whichis a prominent

health care objective in all OECD countries (13).

Notwithstanding the recognized importance of diabetic patients receiving eye screening

services, inequities in such use has been found (14, 15). In particular, lower SES is

consistently associated with underutilization of eye care services, a finding observed

internationally regardless of the type of health care system (16-21). Despite the growing

interest in health inequities in Canada, evidence on inequities in eye care services in the

diabetic population is scarce. Given the increase in the number of low-socioeconomic

families during the past decade and significant clustering of low SES by neighbourhood

(22, 23), it is plausible that SES might play a pivotal role in increasing inequity in

utilization of eye care services in patients with diabetes. It is crucial to understand and to

quantify contributors of inequities in eye care services in order to suggest policy options.

We previously reported the existence of SES-related disparities in eye care services for

Albertans living with diabetes (24), including variations in disparities by geography in the

   

58  

province. The objectives of the study were to identify socio-demographic determinants of

SES-related disparitiesin the use of eye care services in patients with diabetes living in

Alberta; to quantify the contribution of each determinant to the observed disparities; and

to determine whether these might be interpreted as inequities.

3.2. Methods

We used data from the Alberta Diabetes Surveillance System (ADSS) from 1995 to 2009.

The ADSS is the provincial diabetes surveillance system, which provides information on

the incidence, prevalence, and mortality of diabetes and its complications in Alberta,

Canada (25, 26). The ADSS datasets are created by linking and de-identifying data from

several Alberta Health and Wellness (AHW) databases, including discharge abstract

databases, the physician claims databases, the ambulatory care classification system, and

the population registry (25). The ADSS datasets include all diabetic patients, diagnosed

by physicians and covered by the provincial health care system from 1995-2009. The case

definition to identify diabetic patients requires that an individual must have one

hospitalization with an International Classification of Disease, 9th revision (ICD-9) code

of 250, selected from all available diagnostic codes from the hospital discharge abstract

for years 1995-2001, or equivalent ICD-10 codes (E10-E14) for the years after 2001-

2002, or two physician claims with an ICD-9 code of 250 within 2 years, selected from

any of the three available diagnostic codes from the physician claims database (26).

Women with gestational diabetes were excluded from these cases (26). In this study, we

included only diabetic patients over age 20, and individuals whose postal code were

linkable to a specific neighbourhood.

   

59  

We used three different SES indicators: census-based median household income, and the

Canadian material and social deprivation indices (27). The median household income was

obtained from 2006 Canadian census data and linked to the ADSS datasets using

Alberta’s 70 sub-region levels. The Canadian deprivation index, which consists of

separate indices of material and social deprivation, were created based on 2006 Canadian

census data. Both material deprivation index (MDI) and social deprivation index (SDI)

were linked to the ADSS datasets at the sub-region level. The Canadian deprivation index

was developed to measure multiple dimensions of SES and is widely used as a standard

measurement of SES (27). MDI is a composition of educational attainment, employment

status and income level. SDI includes the prevalence of single-parent families as well as

those people living alone, and those who are separated, divorced or widowed.

The use of eye care services by ophthalmologists after a patient’s initial diagnosis of

diabetes was measured by all contacts with ophthalmologists in each year from1995 to

2009. This would include any type of services or procedure provided by ophthalmologists

and claimed for reimbursement by Alberta Health and Wellness [Table S3.1]. In Alberta,

optometrist services were not fully covered by the provincial health care plan during this

time period and were therefore not included in the physician claims database.

To quantify the contribution of an individual determinant to the SES-related Relative

Concentration Index (RCI), we first calculated the RCIs, and 95% confidence intervals,

from 1995 to 2009 based on household income, material and social deprivation

indices.The RCI is commonly used as a standard measurement tool for socioeconomic-

related inequality in the field of health economics and policy (28, 29). The RCI is derived

   

60  

from a concentration curve, where the individuals are ranked by socioeconomic status,

and the cumulative percentage of each group is plotted against the cumulative share of

total health care use (29). We then calculated the RCIs for the population,stratified by age,

sex, Aboriginal status, place of residence (urban/rural), and duration of diabetes.

After obtaining the RCIs for the 15-year period, we applied the decomposition method to

assess the major contributors to socioeconomic-related RCIs. The decomposition of the

RCI was proposed by Wagstaff et al. (30). The basic idea of decomposition is that

disparities in an outcome variable reflect, at a minimum, disparities in various associated

factors (explanatory variables) (30, 31). To apply the decomposition method, age, sex,

duration of diabetes were included as “need” factors (i.e., risk, and thus use, would be

expected to increase with ‘need’ factors), while Aboriginal status, place of residence and

Alberta Health Services (AHS) zones, were included as non-need factors in our analytic

model.In decomposition analysis, the non-need factors are usually considered as

violations of horizontal equity principle (32). Age-sex interactions were generated based

on the identified age-sex related risks of diabetic eye disease (33). The other explanatory

variables included in our analytic models have been well documented as determinants of

health and health care in Canadian literature (23, 34, 35). The decomposition of the RCI

was computed by the following steps. Assuming that the outcome variable for eye care

services by an ophthalmologistis linked to k determinants through a linear regression, it is

possible to decompose the RCI by those k determinants as follows:

𝑅𝐶𝐼 = (𝛽!𝑥!/!

𝜇)𝐶! +𝐺𝐶!𝜇

=  𝐶!   + 𝐺𝐶!/𝜇

Where µμ is the mean of the proportion of eye care services by an ophthalmologist, βk is

the regression coefficients for the explanatory variables, 𝑥! is the mean of individual

   

61  

explanatory variables, 𝐶! is householdincome, MDI, or SDI-related RCI and 𝐺𝐶! is the

generalized concentration index for the error term.

The decomposition analysis was conducted using the following steps:

(1) the outcome variable (i.e., eye care services by an ophthalmologist) was regressed

against the explanatory variables (i.e., age, sex, Aboriginal status, AHS zone and duration

of diabetes) using a probit model. This provides the marginal effects of the explanatory

variables (i.e., βk).

(2) the means of the outcome variable and each of the explanatory variables (i.e., 𝑥!)

were calculated.

(3) the RCI for the outcome variable and the explanatory variables (i.e., C and Ck) were

calculated, using the equation for RCI – 𝐶 = !!!

ℎ!𝑟!   − 1 −  !!

!

!!!as well as the

generalized RCI of the error term (i.e., GC).

The contribution of each determinant to socioeconomic-related RCI in eye care services

was quantified as follows: (1) The absolute contribution of each determinant was

calculated by multiplying its outcome elasticity (βK𝑥!/𝜇) with respect to each

determinant and the relevant RCI of each determinant, (2) The percentage contribution of

each determinant was calculated by dividing its absolute contribution by the RCI of the

outcome variable (36, 37). Positive contribution of each determinant indicates that the

determinant is associated with both socioeconomic status and outcome variables. In other

words, positive contribution of the determinant can be interpreted that the observed

   

62  

disparities can be x % reduced if the determinant is distributed equally across

socioeconomic groups.

In addition to decomposing the RCIs, we calculated the horizontal inequity index (HI),

which can be computed as the RCI minus the sum of the absolute contributions of all

‘need’ factor variables as Wagstaff, et al. proposed (29). HI represents the egalitarian

principle of equal treatment or access for equal need, irrespective of characteristic such as

income, place of residence, etc. A positive (negative) value of HI indicates horizontal

inequity favouring the better-off (worse-off). A zero value of HI means no horizontal

inequity (i.e., eye care is proportionally distributed by need across income, MDI, and SDI

distribution). In decomposition analysis, need factor for eye care services were

represented through age, sex, and duration of diabetes, as these variables reflect

individual’s health care needs (32). All the RCIs, HIs, and decomposition of the RCIs

were estimated using STATA 12 for Microsoft Windows.

3.3. Results

Concentration Indices for each determinant applying different socioeconomic measures

Table 3.1 presents income-related RCIs for the diabetic population stratified by sex, age,

place of residence (i.e., urban and rural), Aboriginal status, and duration of diabetes

during the follow-up period. These RCIs summarize the distribution of eye care services

by the quintiles of SES indicator for each of the explanatory variables [Table S3.2-S3.4].

In both young (less than 65 years of age) and old (more than 65 years of age) groups,

“pro-rich” disparities in eye care services were observed. The income-related RCIs

   

63  

decreased marginally over time, and similarly for both young and old groups. The

magnitude of income-related RCIs for sex and duration of diabetes did not change

substantially over time. It is noteworthy that while “pro-rich”income-related disparities in

urban areas decreased over time, “pro-rich” disparities in rural areas increased during the

same period. The “pro-rich” pattern of RCIs related to non-Aboriginals dropped towards

the null (i.e., no disparity), whiledisparities in Aboriginal population increased, starting

from a “pro-poor” disparity in 1995, becoming increasingly “pro-rich”from 2001 and

thereafter.

Similar to income-related RCIs, MDI-related RCIs have decreasedduring the past 15

years in both male and female, young and old, and shorter and longer duration of diabetes

(Table 3.2). While MDI-related RCIs decreased for both urban and rural areas over time,

the decrease was larger, and became closer to equity in rural compared to urban areas.

Aboriginal status was associated with a change from “pro-rich” to increasingly “pro-

poor”disparities after 2002. In contrast, a constant decrease in “pro-rich” disparities in

non-Aboriginal groups was found over the same time period.

While the income and MDI- related disparities were generally pro-rich, SDI-related RCIs

were generally “pro-poor” (Table 3.3). This was seen for male and female, younger and

older ages, and regardless of diabetes duration. Different patterns were observed,

however, between urban and rural and Aboriginal and non-Aboriginal groups. The urban

area group had consistently “pro-poor” disparities while the rural area group had “pro-

rich” disparities, although over time both pro-poor and pro-rich disparities were reduced

somewhat. The non-Aboriginal population had “pro-poor” disparities, suggesting that

   

64  

individuals living in more socially deprived neighbourhoods were more likely to receive

eye care services, but these disparities decreased over time. In contrast, the SDI-related

RCIs in Aboriginal diabetic individuals were initially pro-poor, but became “pro-rich”

after 1999, suggesting that the eye care services were increasingly concentrated in

Aboriginals living in less socially deprived areas.

Horizontal inequity index (HI)

Over the period from 1995 to 2009, socioeconomic-related horizontal inequitieswere

observed but steadily decreased (Table 3.4-3.6; Figure 3.1). Both income- and MDI-

related HIs were consistently in favour of the rich, suggesting that individuals living in

higher income or less materially deprived neighbourhoods were inclined to use more eye

care services. In contrast, SDI-related HIs were consistently in favour of the poor,

suggesting that those living in more socially deprived neighbourhoods tended to use more

eye care services. In particular, we found that contributions of non-need factors such as

an individual’s SES were always larger than those of need factors in income-, MDI-, and

SDI-related RCIs. As a result, our finding suggests that inequities, among diabetic

patients, in the use of eye care services by an ophthalmologist remain after differences

based on age, sex and duration of diabetes were accounted for.

Decomposition analyses: Contribution of Need, Non-need and Residuals

Table 3.7 shows the contributions for each covariate and the percentage contribution of

each to the overall income-related RCI in the use of eye care services. The largest

contribution to income-related RCI for eye care services comes from MDI. Place of

   

65  

residence - urban and rural areas- was the second largest contribution to the income-

related RCIs over time. Aboriginal status also contributed to “pro-rich” disparities in eye

care services among diabetic patients. Over the 15-year period, the contribution of the

AHS zones to total inequality has fluctuated and, in more recent years, the contribution

became smaller and then became a negative contribution to “pro-rich” disparities (i.e.,

reduction of a pro-rich disparity).

For MDI-related RCIs, the largest contribution was consistently from material

deprivation itself (Table 3.8). The contribution of MDI to total inequality was about 30%

in 1995, but this increased to 67% by 2005 and remained the largest contribution to the

MDI-related RCI. Place of residence - urban versusrural areas- wasthe second largest

contribution according to the MDI-related RCI decomposition analyses. Unlike the

negative contribution to income-related RCIs, the combination of need factors (i.e., age-

sex interaction and duration of diabetes) showed a small positive value, indicating a

relatively small contribution to a “pro-rich” direction, except for 2009 when the

contribution was to a “pro-poor” direction. In relation to the AHS zones, the “pro-rich”

contributions to the MDI-related RCI were found although the magnitude of the

contributions became smaller in 2007 and for every year after it showed negative

contributions to the MDI-related RCIs.

Table 3.9 demonstrates contributions for each covariate to SDI-related RCI in the use of

eye care services over the 15-year period. In 2009, 68% of SDI-related RCI in the use of

eye care services was explained by social deprivation itself. Place of residence –urban or

rural areas- contributed from 27% to 36% of the “pro-poor” direction over the 15 year

   

66  

period. In addition, need factors such as age-sex interaction and duration of diabetes

contributed to “pro-poor” RCIs over the same period. It is noteworthy that AHS zones

contributed to “pro-poor” RCIs from 1995 to 2003, with a decreasing contribution to the

total RCIs. After 2004, the contribution of AHS zones became a negative value,

indicating that the contributions of AHS zones contributed to the “pro-rich” direction of

SDI-related RCIs.

3.4. Discussion

Major contributors to existing socioeconomic inequities

While the contribution of each determinant varies, we observed two major factors that

caused inequities across income- and MDI-related inequities – material deprivation and

urban and rural place of residence. For both income and MDI-related inequities, material

deprivation was unsurprisingly found to be the most important contributor to the “pro-

rich” inequity – accounting for about 25 to 124%of absolute percentage of contribution to

the measured RCIs over the follow-up period. This finding suggests that material

deprivation is responsible for presenting important barriers to use of eye care services

among diabetic patients under Canada’s publicly funded health care system. This finding

is consistent with previous research on eye screening services in patients with diabetes in

the U.K and Australia. These studies indicate that living in deprived area is known to be

associated with substantial variation in the uptake of screening service among diabetic

patients (16, 18, 20).

   

67  

The lesser contribution of income to pro-rich inequity in the use of eye care services may

be explained by the fact that the aim of the Canadian healthcare system to provide

healthcare services regardless of ability to pay works well in the case of provision of eye

care services by an ophthalmologist among diabetic patients in Alberta. However, the

contribution of income to overall pro-rich inequity cannot be overlooked. In fact, income

was observed to be the largest contributor in both 2001 and 2002 when contributions of

material deprivation were smaller than those of income. Several studies have revealed

inequities in specialist care, suggesting a trend that favours higher-income individuals in

specialist care services under the publicly funded health care system (9, 13, 22). Both

higher educational attainment and higher income have been associated with higher rates

of initial appointments for specialist care (9, 38). The contribution of place of residence to

economic-related RCIs, appears to reflect, at least in part, the fact that urban residence is

closely correlated with higher income; people living in urban areas tend to have both

higher income and education in addition to secure employment status (17). In fact,

according to our decomposition analysis of both income- and MDI-related RCIs, lower

income and more materially deprived individuals are concentrated in rural areas. This

may explain why place of residence has appeared as major contributor to economic-

related RCIs. At the same time, the distribution of ophthalmology services is more

concentrated in urban areas, facilitating easy access to the services for more affluent

urban dwellers (39).

Social deprivation-related inequities were largely explained by social deprivation itself

over time and this finding suggests that social deprivation needs to be main interest of

health care policy in order to reduce social deprivation-related inequities in the use of eye

care services among diabetic patients. The potential explanations for “pro-poor”

68

inequities related to social-related RCIs are several. “Pro-poor” inequities may be

explained by the fact that most urban areas in Alberta have lower SDI, which tells us that

urban dwellers are more likely to live in areas that while more “socially deprived” have

greater availability of ophthalmology services. In fact, the SDI-related RCIs, stratified by

place of residence, indicate that the RCIs in urban areas were consistently “pro-poor”

while those in rural area were “pro-rich”. Due to the complexity of the concept of social

capital, it has been suggested that not all component of social capital may be related to

use of health care services (40). To understand the association between social deprivation

and the use of eye care services, further investigation is needed.

Aboriginal status also emerged as a major contribution to the “pro-rich” inequity in the

use of eye care services. Access to health care is an issue for Aboriginal populations as

they are less likely than the overall population to regularly visit a physician, and are more

likely to report having unmet health care needs (41). For example, the vast majority of

First Nations people with diabetes who live on reserves in British Columbia do not have

access to annual examinations by an eye care professional (42). This fact might influence

our interpretation of the results that Aboriginal populations encounter difficulties in

receiving eye care services. Given the fact that a large proportion of the Aboriginal

populations is concentrated in the AHS North zone, and live in rural areas, they might

face more complex barriers to the use of eye care services.

Concern with inequities in eye care services related to place of residence and Aboriginal

populations have been previously been identified in Alberta (43). To overcome

geographic barriers to eye care services, the Department of Ophthalmology at the

69

University of Alberta developed a tele-ophthalmology program in 1998 (43). Alberta’s

comprehensive tele-ophthalmology program enables specialists to provide standard eye

care services and diagnose ocular complications. The implementation of Alberta’s tele-

ophthalmology program may have influenced changes in direction and magnitude of

inequities observed after 1998 in rural areas and among Aboriginal individuals. For

example, income-related RCI were “pro-poor” in the Aboriginal population before 1998,

but thereafter changed to “pro-rich” disparities. In addition, the magnitude of income-

related RCI becomes slightly larger after 1998. Similar patterns were found in the

direction of MDI-related RCI, which also reversed in the Aboriginal group after 1998. In

addition, the magnitude of MDI-related RCI became smaller. This finding may suggest

that the implementation of Alberta’s tele-ophthalmology produced unintended inequities

in rural areas and among Aboriginal populations in the province as a whole, while

providing a better access to eye care services to individuals living in small and remote

community in the AHS north zone.

In addition, residence in a specific AHS zone appeared to be a major factor contributing

to “pro-rich” patterns of income- and MDI-related RCIs and to the “pro-poor” pattern of

SDI-related RCI. In particular, we observed “pro-poor” contributions to SDI-related RCI

until 2003, but, beginning in 2004, area of residence became a negative contributor -

moving inequities to become “pro-rich” in 2004, and the magnitudes of the positive

contributions became larger as one moved closer to 2009. It is possible that this trend

might be associated with the regionalization implemented as a strategy for Alberta’s

healthcare system improvement. Prior to establishment of the current 5 AHS zones,

health services in Alberta have undergone several governance re-organizations. Until

March 2009, Alberta had 9 health regions (Chinook, Palliser, Calgary, David Thompson,

70

East Central, Capital, Aspen, Peace Country, and Northern lights). Before 2003, there

were 17 regional health authorities and 2 provincial boards for cancer and mental health

(44). Reforms in the structure of health regions in the past were seen as opportunities to

provide integrated health care service in addition to changes in governance and

management systems (44). Thus, changes in the distribution of eye care services during

our study period may reflect differences in these governance and management systems.

Finally, unlike economic-related RCIs, need factors consistently contributed to an

increase in “pro-poor” direction of social-related RCIs. This may suggest that age, sex,

and duration of diabetes must be considered if one wishes to reduce SDI-related

inequities. In particular, individuals who are female and older than 65 years of age tend to

contribute more to the pro-poor SDI-related RCIs. It is important to highlight that the

contributions of need factors to the SDI-related RCIs increased over time and reached

approximately 32% of contributions by 2009.

Limitations

While decomposition of the RCI allows one to identify the factors contributing most to

observed inequities in health care, it should be recognized that the decomposition analysis

is not able to provide causal pathways between socio-demographic determinants and

health care use. The contribution of a determinant to the observed inequity usually

depends on the products of its estimated elasticity with the health care outcomes at the

sample means and the RCI of the factor itself, so it may fail to explain causal pathways

(45). In addition, our socioeconomic indicator was at the level of Alberta’s 70 sub-

regions, so it may be unable to fully explain the variation of socioeconomic status with

71

the same geographical region in the use of eye screening services, but at least it is

possible to draw a conclusion that the variation between different sub-regions where there

are different socioeconomic characteristics.

The decomposition method is a deterministic approach and there could be other factors

(e.g., clinical, cultural and health care system related determinants not included in the

model) that have may contributed to the inequities in eye care services. While we

included age, sex, and duration of diabetes as ‘need’ factors for eye care services, there

are many other important clinical indicators of need for eye care services. Unfortunately,

our data were limited to the administrative databases, which lack information on these

important clinical parameters. In addition, decomposing inequity based on need-adjusted

utilization has been shown to suffer from several conceptual and measurement-related

limitations (40). For example, our decomposition model assumes homogeneity in

behaviours across socioeconomic groups of populations. Lastly, as provincial

administrative data only included ophthalmology services, patients who used optometry

services were unintentionally excluded in our analyses. As of the end of 2007, eye

examinations by optometrists were covered by the provincial healthcare plan for patients

with diabetes (3). Considering the large number of optometrists in Alberta (46), further

examination of potential inequities, including information on services by both

ophthalmologists and optometrists, is required.

3.5. Conclusion

Despite clear guidelines intended to address early detection and timely treatment of DR

for all patients with diabetes, our findings suggest that individuals living with diabetes in

72

Alberta experience inequities in the use of eye care service based on socio-demographic

characteristics unrelated to clinical need. In particular, material deprivation, urban/rural

residence and AHS zone of residence appear to be the main contributors to the existing

economic-related inequities in the use of eye care services. For social-related inequities,

social deprivation, place of residence and AHS zones are the main contributors to existing

SDI-related inequities, in addition to “need factors” of age, sex and duration of diabetes.

In order to provide more equitable eye care services for all diabetic patients, the factors

identified in this study should be further investigated, and if necessary, targeted by policy

interventions at community, regional, and provincial levels.

   

73  

Young(<65)Old(>=$65)

Male

Female

Urban

Rural

AboriginalNon9aboriginalShorter

Longer

1995

0.033

0.037

0.038

0.025

0.021

0.002

90.040

0.030

0.032

0.035

1997

0.017

0.041

0.037

0.018

0.016

0.013

90.033

0.026

0.034

0.016

1999

0.024

0.044

0.043

0.020

0.020

0.018

90.023

0.029

0.038

0.021

2001

0.023

0.036

0.033

0.024

0.016

0.031

0.014

0.025

0.030

0.028

2003

0.016

0.041

0.032

0.021

0.012

0.022

0.023

0.024

0.029

0.023

2005

0.015

0.035

0.028

0.018

0.010

0.025

0.028

0.021

0.026

0.019

2007

0.031

0.035

0.037

0.024

0.018

0.033

0.038

0.027

0.033

0.029

2009

0.023

0.025

0.027

0.017

0.010

0.033

0.033

0.018

0.022

0.025

DurationHofHdiabetes

Sex

PlaceHofHresidence

AboriginalHstatus

Year

Age

Young(<65)

Old(>=$65)

Male

Female

Urban

Rural

Aboriginal

Non9aboriginalShorter

Longer

1995

0.087

0.086

0.088

0.090

0.055

0.056

0.053

0.082

0.091

0.083

1997

0.064

0.081

0.080

0.074

0.051

0.058

0.067

0.070

0.083

0.064

1999

0.062

0.077

0.077

0.067

0.048

0.051

0.038

0.065

0.082

0.054

2001

0.055

0.064

0.063

0.063

0.040

0.050

0.005

0.056

0.068

0.055

2003

0.046

0.066

0.061

0.057

0.038

0.024

90.007

0.055

0.064

0.050

2005

0.040

0.060

0.056

0.050

0.037

0.022

90.041

0.052

0.057

0.046

2007

0.048

0.052

0.055

0.052

0.040

0.010

90.044

0.050

0.056

0.049

2009

0.034

0.045

0.044

0.042

0.031

0.007

90.035

0.040

0.043

0.043

AboriginalGstatus

Year

Age

PlaceGofGresidence

DurationGofGdiabetes

Sex

Tab

le 3

.1. I

ncom

e-re

late

d R

CI b

y ag

e, s

ex, p

lace

of r

esid

ence

, Abo

rigin

al st

atus

, and

dur

atio

n of

dia

bete

s

Tab

le 3

.2. M

DI-

rela

ted  RCI  by  age,  sex,  place  of  residence,  Aboriginal  status  and  duration  of  diabetes  

Tables and Figures

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

74  

Income

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

Need3

!0.005

!0.005

!0.003

!0.004

!0.004

!0.004

!0.002

!0.003

!0.003

!0.003

!0.003

!0.004

!0.004

!0.004

!0.005

Non4need

0.026

0.030

0.023

0.029

0.027

0.023

0.025

0.022

0.021

0.020

0.017

0.019

0.024

0.024

0.019

Total3inequality

0.031

0.036

0.027

0.033

0.031

0.027

0.028

0.027

0.026

0.026

0.023

0.024

0.030

0.027

0.021

Income

3HI

0.036

0.041

0.030

0.037

0.035

0.031

0.031

0.030

0.029

0.029

0.026

0.029

0.034

0.031

0.026

Tab

le 3

.3. S

DI-

rela

ted

RC

I by

age,

sex

, pla

ce o

f res

iden

ce, A

borig

inal

sta

tus,

and

dur

atio

n of

dia

bete

s

Tab

le 3

.4. I

ncom

e-re

late

d H

Is in

the

use

of e

ye c

are

serv

ices

am

ong

diab

etic

pat

ient

s ove

r a 1

5-ye

ar p

erio

d

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Young(<65)Old(>=$65)

Male

Female

Urban

Rural

AboriginalNon9aboriginalShorter

Longer

1995

90.050

90.063

90.053

90.077

90.034

0.034

90.060

90.062

90.069

90.048

1997

90.037

90.053

90.040

90.062

90.028

0.029

90.067

90.047

90.050

90.049

1999

90.026

90.043

90.029

90.054

90.023

0.050

0.007

90.039

90.041

90.038

2001

90.021

90.039

90.027

90.043

90.016

0.037

0.017

90.033

90.034

90.032

2003

90.021

90.035

90.027

90.042

90.018

0.025

0.015

90.034

90.035

90.032

2005

90.023

90.035

90.032

90.040

90.023

0.006

0.030

90.037

90.034

90.036

2007

90.019

90.028

90.022

90.039

90.020

0.017

0.029

90.031

90.032

90.024

2009

90.014

90.027

90.021

90.034

90.020

0.023

0.033

90.028

90.029

90.020

DurationHofHdiabetes

Year

Age

Sex

PlaceHofHresidence

AboriginalHstatus

   

75  

Tab

le 3

.5. M

DI-

rela

ted

HIs

in th

e us

e of

eye

car

e se

rvic

es a

mon

g di

abet

ic p

atie

nts o

ver a

15-

year

per

iod

Tab

le 3

.6. S

DI-

rela

ted

HIs

in th

e us

e of

eye

car

e se

rvic

es a

mon

g di

abet

ic p

atie

nts o

ver a

15-

year

per

iod

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

MDI

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

Need

0.003

0.002

0.003

0.002

0.002

0.002

0.003

0.002

0.002

0.002

0.002

0.002

0.002

0.002

0.002

Non3need

0.070

0.070

0.060

0.063

0.057

0.055

0.051

0.052

0.050

0.046

0.047

0.044

0.047

0.043

0.039

Total8inequality

0.089

0.087

0.077

0.077

0.072

0.068

0.063

0.061

0.059

0.053

0.053

0.049

0.053

0.048

0.043

MDI8HI

0.086

0.085

0.073

0.075

0.070

0.066

0.060

0.059

0.057

0.051

0.051

0.047

0.051

0.047

0.041

SDI

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

Need

!0.010

!0.009

!0.007

!0.008

!0.008

!0.007

!0.006

!0.006

!0.007

!0.007

!0.008

!0.009

!0.009

!0.009

!0.009

Non3need

!0.046

!0.040

!0.035

!0.029

!0.023

!0.031

!0.021

!0.031

!0.029

!0.027

!0.032

!0.027

!0.025

!0.018

!0.020

Total8ineqaulity

!0.066

!0.056

!0.051

!0.048

!0.041

!0.048

!0.035

!0.036

!0.035

!0.032

!0.036

!0.032

!0.031

!0.025

!0.028

SDI8HI

30.05

630.04

730.04

430.04

130.03

430.04

030.02

930.03

030.02

830.02

430.02

830.02

430.02

230.01

730.01

9

   

76  

Tab

le 3

.7. D

ecom

posi

tion

of in

com

e-re

late

d R

CI,

1995

-200

9

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Age$sex2

Age$sex3

Age$sex4

SubAge$sex

Duration4of4DM

Need

Income

MDI

SDI

Residence

Aboriginal4Caglary

CentralEdmontonNorth

Sub$AHS4Non$need

1995

%4of4contribution

$0.3

$3.1

$9.0

$12.4

$4.6

$17.0

$8.1

60.3

$3.9

21.1

5.0

$45.0

29.7

23.5

1.7

9.9

84.4

1996

%4of4contribution

$0.1

$2.1

$7.2

$9.4

$3.5

$12.9

$0.8

45.9

$1.4

17.4

4.3

$25.9

24.7

17.2

2.1

18.1

83.5

1997

%4of4contribution

0.2

$0.4

$5.6

$5.7

$5.7

$11.4

$4.3

49.6

$2.2

21.0

5.9

$43.8

31.9

23.9

1.0

12.9

82.9

1998

%4of4contribution

$0.1

$1.3

$7.2

$8.5

$4.2

$12.7

22.9

42.8

0.8

17.5

6.3

$51.5

29.5

20.7

0.7

$0.5

89.8

1999

%4of4contribution

0.0

$0.5

$8.0

$8.6

$4.5

$13.0

16.1

43.1

4.6

18.8

6.6

$57.5

30.3

23.4

0.6

$3.2

86.0

2000

%4of4contribution

0.0

$0.5

$10.1

$10.6

$4.1

$14.7

36.6

43.3

$12.0

21.2

8.2

$73.8

32.4

29.5

$0.7

$12.7

84.6

2001

%4of4contribution

0.3

1.2

$5.7

$4.3

$4.1

$8.4

41.9

26.0

$3.6

19.7

8.3

$61.8

32.3

25.6

$1.1

$5.0

87.2

2002

%4of4contribution

0.1

0.2

$7.9

$7.7

$3.0

$10.7

44.6

32.8

$26.7

20.6

4.0

$47.4

31.5

23.7

$1.9

6.1

81.3

2003

%4of4contribution

$0.1

0.5

$9.1

$8.7

$3.0

$11.7

23.6

31.4

$24.9

25.9

4.9

$34.1

31.9

24.5

$2.1

20.2

81.0

2004

%4of4contribution

$0.3

0.5

$8.9

$8.7

$3.0

$11.8

10.5

61.8

$30.4

21.7

2.2

$52.0

33.9

29.8

$2.1

9.6

75.5

2005

%4of4contribution

$0.3

$0.1

$11.1

$11.5

$3.8

$15.3

$46.2

124.4

$39.2

22.6

3.9

$55.9

35.0

30.5

$2.3

7.3

72.8

2006

%4of4contribution

$0.4

$0.5

$11.1

$12.1

$4.5

$16.6

$1.0

91.1

$40.4

21.1

4.7

$73.8

39.5

37.0

$2.0

0.6

76.0

2007

%4of4contribution

$0.3

$0.5

$9.4

$10.2

$4.4

$14.5

8.2

79.3

$33.5

19.5

5.8

$64.7

34.6

31.6

$1.8

$0.3

79.0

2008

%4of4contribution

$0.5

0.1

$10.3

$10.7

$5.3

$15.9

11.6

74.8

$27.1

23.6

7.6

$80.4

41.0

39.2

$1.8

$2.0

88.5

2009

%4of4contribution

$0.7

$0.2

$13.0

$14.0

$7.7

$21.7

9.0

86.4

$39.8

29.6

9.6

$101.8

52.5

48.8

$3.8

$4.2

90.5

*Age

-sex2

Age

-sex3

Age

-sex4

Non$need'facto

rs

Youn

ger (

<65)

& F

emal

e

Old

er(>

=65)

&Fe

mal

eO

lder

(>=6

5)&

Mal

e

Year

Decomposition

Need'facto

rs

   

77  

Tab

le 3

.8. D

ecom

posi

tion

of M

DI-

rela

ted

RC

I, 19

95-2

009

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Age$sex2Age$sex3Age$sex4

SubAge$sex

Duration4of4DM

Need

Income

MDI

SDI

Residence

Aboriginal4Caglary

Central

EdmontonNorth

Sub$AHS4

Non$need

1995

%4of4contribution

0.5

1.1

1.4

3.1

$0.1

2.9

$2.3

29.2

0.8

22.3

4.8

$20.0

19.7

$10.6

35.1

24.2

79.0

1996

%4of4contribution

0.4

1.2

1.2

2.8

0.0

0.2

$0.3

26.5

0.4

20.2

4.9

$13.8

19.4

$9.6

32.9

29.0

80.7

1997

%4of4contribution

0.5

2.2

2.2

4.9

$0.6

4.4

$1.3

24.3

0.4

19.6

5.5

$19.9

21.4

$10.7

38.7

29.4

78.0

1998

%4of4contribution

0.3

1.6

0.8

2.6

$0.4

2.2

8.1

25.2

$0.1

19.2

6.9

$27.8

23.2

$11.2

38.3

22.5

81.7

1999

%4of4contribution

0.3

1.8

0.8

2.9

$0.6

2.3

5.8

25.7

$0.3

20.8

7.0

$31.4

24.5

$13.0

40.1

20.3

79.4

2000

%4of4contribution

$0.1

2.1

0.9

2.9

$0.5

2.3

12.6

23.9

0.4

21.8

8.0

$37.3

24.3

$14.1

41.5

14.4

81.1

2001

%4of4contribution

0.4

2.6

2.0

4.9

$0.7

4.2

16.2

16.1

0.1

21.8

8.7

$34.7

27.4

$15.4

42.0

19.3

82.1

2002

%4of4contribution

0.2

2.4

1.0

3.6

$0.3

3.3

17.2

20.3

$0.2

19.6

4.2

$26.4

26.6

$14.1

38.1

24.2

85.4

2003

%4of4contribution

$0.2

2.9

1.4

4.1

$0.3

3.8

9.0

19.1

$0.4

24.1

5.1

$18.6

26.2

$14.0

34.7

28.2

85.2

2004

%4of4contribution

$0.7

3.3

1.9

4.5

$0.2

4.3

4.4

38.0

$0.9

21.8

2.6

$30.8

29.7

$17.2

38.1

19.8

85.8

2005

%4of4contribution

$0.7

3.1

2.1

4.5

$0.3

4.2

$17.5

67.6

$1.5

20.0

4.1

$29.2

26.9

$14.9

33.3

16.1

88.7

2006

%4of4contribution

$1.0

3.2

2.2

4.4

$0.2

4.2

$0.4

56.8

$2.1

21.4

5.5

$44.1

34.8

$19.3

36.9

8.2

89.4

2007

%4of4contribution

$0.9

3.2

1.7

4.0

$0.6

3.5

3.7

56.2

$2.3

21.8

6.7

$43.6

34.2

$18.4

31.1

3.1

89.3

2008

%4of4contribution

$1.0

3.4

1.8

4.3

$0.8

3.5

5.2

52.1

$1.3

25.9

8.6

$53.1

39.5

$21.4

34.8

$0.2

90.2

2009

%4of4contribution

$1.2

3.8

2.2

4.7

$1.1

$0.6

3.6

53.4

$2.0

28.5

9.6

$59.7

44.2

$22.7

35.6

$2.5

90.5

*Age

-sex2

Age

-sex3

Age

-sex4

Old

er(>

=65)

&M

aleO

lder

(>=6

5)&

Fem

ale

Non$need'facto

rsYear

Decomp

osition

Need'facto

rs

Youn

ger (

<65)

& F

emale

   

78  

Tab

le 3

.9. D

ecom

posi

tion

of S

DI-

rela

ted

RC

I, 19

95-2

009

Age$sex2Age$sex3Age$sex4

Sub-A

ge$sex

Duration

-of-DM

Need

Incom

eMD

ISDI

Residence

Aboriginal-Caglary

Central

Edmo

ntonNorth

Sub$A

HS-N

on$ne

ed

1995

%-of-contribu

tion

1.13.3

8.412.8

2.315.1

0.89.5

7.232.9

3.1$3.8

16.9

$31.7

35.9

17.3

70.8

1996

%-of-contribu

tion

0.83.8

9.013.6

2.913.6

0.19.6

3.732.7

3.6$2.8

18.4

$32.7

38.2

21.2

70.9

1997

%-of-contribu

tion

0.62.9

7.010.5

3.113.6

0.48.1

4.529.2

3.6$3.1

18.4

$36.0

43.0

22.3

68.2

1998

%-of-contribu

tion

0.54.0

8.412.9

3.015.8

$3.2

8.9$1.9

29.7

4.7$3.6

19.6

$39.7

44.3

20.7

58.9

1999

%-of-contribu

tion

0.54.1

10.3

14.9

3.518.4

$2.6

9.2$9.6

34.4

5.0$2.3

21.0

$50.0

49.9

18.6

55.0

2000

%-of-contribu

tion

$0.1

3.810.0

13.7

1.815.5

$4.6

6.418.3

29.0

4.3$0.3

15.6

$46.1

42.1

11.2

64.6

2001

%-of-contribu

tion

0.33.1

9.513.0

3.216.2

$7.7

5.17.5

35.4

5.32.3

19.7

$59.9

52.7

14.8

60.5

2002

%-of-contribu

tion

0.34.4

10.8

15.4

2.618.0

$8.1

5.646.9

27.0

2.43.5

15.6

$51.6

43.8

11.4

85.2

2003

%-of-contribu

tion

$0.3

5.213.1

18.1

2.720.8

$4.4

4.943.0

32.2

2.84.0

13.4

$52.9

39.4

3.982.4

2004

%-of-contribu

tion

$0.8

6.314.6

20.0

3.323.4

$2.2

5.055.7

28.0

1.48.9

12.6

$67.6

41.7

$4.4

83.6

2005

%-of-contribu

tion

$0.8

5.414.0

18.6

3.522.1

7.56.4

54.4

21.9

1.88.8

8.5$52

.531.4

$3.9

88.1

2006

%-of-contribu

tion

$1.2

6.116.8

21.8

5.227.0

0.34.6

66.4

24.1

2.416.5

9.5$74

.334.8

$13.5

84.3

2007

%-of-contribu

tion

$1.2

6.417.4

22.7

5.728.4

$2.9

4.470.0

26.9

2.920.6

8.7$81

.032.4

$19.3

82.1

2008

%-of-contribu

tion

$1.1

6.421.5

26.8

7.534.2

$4.5

3.365.5

34.3

3.832.5

7.4$10

8.538.8

$29.8

72.6

2009

%-of-contribu

tion

$1.1

5.520.3

24.6

7.532.1

$2.6

1.768.1

29.6

3.432.4

3.3$95

.031.7

$27.6

72.6

*Age

-sex2

Age

-sex3

Age

-sex4

Non$need'facto

rs

Youn

ger (

<65)

& F

emale

Old

er(>

=65)

&M

aleO

lder

(>=6

5)&

Fem

ale

Year

Decomp

osition

Need'facto

rs

   

79  

Figure 3.1. The trends of horizontal inequity indices (HIs) over a 15-year period, 1995-2009

!0.080%

!0.060%

!0.040%

!0.020%

0.000%

0.020%

0.040%

0.060%

0.080%

0.100%

1995% 1996% 1997% 1998% 1999% 2000% 2001% 2002% 2003% 2004% 2005% 2006% 2007% 2008% 2009%Income'HI' MDI'HI' SDI'HI'

   

80  

References

1. Hooper P, Boucher MC, Cruess A, Dawson KG, Delpero W, Greve M, et al. Canadian Ophthalmological Society Evidence-Based Clinical Practice Guidelines for the Management of Diabetic Retinopathy. Can J Ophthalmol. 2012;47(2): 91-101. 2. Browing D. Chapter 14. Diabetic Retinopathy. In: Browing D, editor. Diabetic Retinopathy: Evidence-Based Management: Springer; 2010. p. 369-85. 3. Rudnisky CJ, Tennat MTS, Johnson JA, Balko SU. Chapter 9 : Diabetes and Eye Disease in Alberta. In: Johnson JA,editor Alberta Diabetes Atlas 2011. Edmonton: Institute of Health Economics 2011. p.153-74. 4. Buhrmann R, Hodge W, Beardmore J, Baker G, Lowcock B, Pan I, Bovell A. Section 2. The Epidemiologoy of Blindness and Visual Imparment in Canada : Diabetic Retinopathy. Foundations for a Canadian Vision Health Strategy : Towards Preventing Avoidable Blindness and Promoting Vision Health. Toronto:The National Coalition for Vision Health; 2007. p.12-26. 5. Perruccio AV, Badley EM, Trope GE. A Canadian population-based study of vision problems: assessing the significance of socioeconomic status. Can J Ophthalmol. 2010;45(5):477-83. 6. Klein K, Eden B. Overview of Epidemiologic Studies of Diabetic Retinopathy. Ophthalmic Epidemiol. 2007;14(4):179-83. 7. Bond SR, Advani A, Altomare F, Stockl F. Canadian Diabetes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada: Retinopathy. Can J Diabetes. 2013;37(suppl1):S137-S41. 8. Sanmartin C, Gilmore J. Diabetes care in Canada: Results from selected provinces and territory, 2005. Ottawa: Statistics Canada, 2006. 9. Sara A. Does Equity in Healthcare Use Vary across Canadian Provinces? Healthcare Policy. 2008;3(4):83-99. 10. Romanow RJ. Building on Values: The Future of Health Care in Canada. Ottawa: Commission on the Future of Health Care in Canada, 2002. 11. Kozousek V, Brown MG, Cottle R, Hicks VA, Langille DB, Dingle J. Use of ophthalmologic services by diabetic patients in Nova Scotia. Can J Ophthalmol. 1993 28(1):7-10. 12. Koenekoop RK, Gomolin JE. The management of age-related macular degeneration: patterns of referral and compliance in seeking low-vision aids. Can J Ophthalmology. 1995;30(4):208-10. 13. van Doorslaer E. Masseria C. Income-Related Inequality in the Use of Medical Care in 21 OECD Countries. OECD health working paper. Paris: Organization for Economic Co-operation and Development. May, 2004.

   

81  

14. Moss SE, Klein R, Klein BEK. Factors Associated With Having Eye Examinations in Persons With Diabetes. Arch Fam Med. 1995;4(6):529-34. 15. Saadine J, Fong D, Yao J. Factors Associated With Follow-Up Eye Examinations Among Persons With Diabetes. Retina. 2008;28(2):195-2001. 16. Kliner M, Fell G, Gibbons C, Dhothar M, Mookhtiar M, Cassels-Brown A. Diabetic retinopathy equity profile in a multi-ethnic, deprived population in Northern England. Eye (Lond). 2012;26(5):671-7. 17. Keeffe JE, Weih LM, McCarty CA, Taylor HR. Utilisation of eye care services by urban and rural Australians. Brit J Ophthalmol. 2002;86(1):24-7. 18. Fraser S, Edwards L. Health Equity Audit - Diabetic Retinopathy Screening : Southampton City, Hampshire, Isle of Wight and Portsmouth City PCTs. Southampton City PCT, 2010 July. 19. McCarty CA, Lloyd-Smith CW, Lee SE, Livingston PM, Stanislavsky YL, Taylor HR. Use of eye care services by people with diabetes: the Melbourne Visual Impairment Project. Brit J Ophthalmol. 1998;82(4):410-4. 20. Scanlon PH, Carter SC, Foy C, Husband RF, Abbas J, Bachmann MO. Diabetic retinopathy and socioeconomic deprivation in Gloucestershire. J Med Screen. 2008;15(3):118-21. 21. Millett C, Dodhia H. Diabetes retinopathy screening: audit of equity in participation and selected outcomes in South East London. J Med Screening. 2006;13(3):152-5. 22. Veugelers PJ, Yip AM. Socioeconomic disparities in health care use: Does universal coverage reduce inequalities in health? J Epidemiol Community Health. 2003;57(6):424-8. 23. Lemstra M, Neudorf C, Opondo J. Health disparity by neighbourhood income. Can J Public Health. 2006;97(6):435-9. 24. Hwang J, Bowen S, Rudnisky C, Johnson J. Socioeconomic Disparities in Ophthalmology Services among Diabetic Patients in Alberta. Can J Diabetes. 2012;36(5, Supplement):S73. 25. Alberta Health and Wellness. Alberta Diabetes Strategy, 2003-2013: Alberta Health and Wellness; 2003. 26. Johnson JA, Balko SU, Hugel G. Chapter 1 : Background and methods. In: Johnson JA, ed. Alberta Diabetes Atlas 2011. Edmonton: Institute of Health Economics; 2011. p. 3-12. 27. Pampalon R, Hamel D, Gamache P, Raymond G. A deprivation index for health planning in Canada. Chronic Diseases In Canada. 2009;29(4):178-91. 28. O'Donnell O, van Dooslaer E, Wagstaff A, & Lindelow M. Chapter 8. The Concentration Index. Analyzing Health Equity Using Household Survey Data : A

   

82  

Guide to Techniques and Their Impementation. Washington, DC: The World Bank; 2008. p. 95-108. 29. Haper S, Lynch J. Methods for Measuring Cancer Disparities : Using Data Relevant to Healthy people 2010 Cancer-Related Objectives. Bethesda, MD: National Cancer Institute; 2005. 30. Wagstaff A, van Doorslaer E, Watanabe N. On decomposing the causes of health sector inequalities with an application to malnutrition inequalities in Vietnam. J Econometrics. 2003;112(1):207-23. 31. O'Donnell OvD, E.; Wagstaff, A.; Lindelow, M. Chapter 13. Explanning Socioeconomic-Related Health Inequality : Decomposition of the Concentration Index. Analyzing Health Equity Using Household Survey Data : A Guide to Techniques and Their Impementation. Washington, DC: The World Bank; 2008. p. 159-64. 32. Watanabe R, Hashimoto H. Horizontal inequity in healthcare access under the universal coverage in Japan; 1986–2007. Soc Sci Med. 2012;75(8):1372-8. 33. Kostev K, Rathmann W. Diabetic retinopathy at diagnosis of type 2 diabetes in the UK: a database analysis. Diabetologia. 2013;56(1):109-11. 34. Frohlich KL, Ross N, Richmond C. Health disparities in Canada today: some evidence and a theoretical framework. Health Policy. 2006;79(2-3):132-43. 35. Raphael D. Health inequalities in Canada: current discourses and implications for public health action. Critical Public Health. 2000;10(2):193-216. 36. Yiengprugsawan V, Lim L, Carmichael G, Dear K, Sleigh A. Decomposing socioeconomic inequality for binary health outcomes: an improved estimation that does not vary by choice of reference group. BMC Res Notes. 2010;3(1):57. 37. Hosseinpoor AR, Van Doorslaer E, Speybroeck N, Naghavi M, Mohammad K, Majdzadeh R, et al. Decomposing socioeconomic inequality in infant mortality in Iran. Int J Epidemiol. 2006;35(5):1211-9. 38. The Nuffield Trust. Equity in Health and Healthcare: The Nuffield Trust for Research and Policy Studeise in Health Services. The Nuffield Trust; 2003. 39. Abu-Zaineh M, Mataria A, Moatti J-P, Ventelou B. Measuring and decomposing socioeconomic inequality in healthcare delivery: A microsimulation approach with application to the Palestinian conflict-affected fragile setting. Soc Sci Med. 2011;72(2):133-41. 40. Perry M, Williams RL, Wallerstein N, Waitzkin H. Social Capital and Health Care Experiences Among Low-Income Individuals. Am J Public Health. 2008;98(2):330-6. 41. The Chief Public Health Officer. Report on the state of public health in Canada: addressing health inequalities. 2008.

83

42. Kaur H, Maberley D, Chang A, Hay D. The current status of diabetes care, diabetic

retinopathy screening and eye-care in British Columbia's First Nations

Communities, Int J Cicumpolar Health. 2004;63(3):277-85.

43. Ng M, Nathoo N, Rudnisky CJ, Tennant MT. Improving access to eye care:

teleophthalmology in Alberta, Canada. J Diabetes Sci Technol. 2009;3(2):289-96.

44. Golden-biddle K, Hinings CR, Casebeer A, Pablo A, Reay P. Organizational Change

in Healthcare with Special Reference to Alberta. Ottawa:Candian Health Services

Research Foundation;2006.

45. Knowles JC, Bales, S, Cuong LQ, Oanh TT, Luong DH. Health equity in Viet Nam a

situational analysis focused on maternal and child mortality. Vietnam:United for

Childern;2010.

46. Jin Y-P, Trope GE. Eye care utilization in Canada: disparity in the publicly funded

health care system. Can J Ophthalmology. 2011;46(2):133-8.

   

84  

Supplementary Tables

Table S3.1. Alberta physician claims data

Procedure   Code   Description  

Retinal  Laser  Treatment  (Retinal  Photocoagulation)   28.5A   Focal  and/or  pan  retinal  

photocoagulation  

Vitrectomy  Surgery   28.72A   Vitreous  cavity  washout  

  28.72B   Total  vitrectomy  

  28.74A   Discission  of  vitreous/retinal  adhesions  

  28.74B  Stripping  of  premacular  membrane,  Associated  vitrectomy  and  retinal  encircling  

Cataract  Surgery   27.72   Insertion  of  intraocular  lens  prosthesis  with  cataract  extraction,  one-­‐stage  

 Laser  Treatment  for  Glaucoma  

26.52  (1995-­‐2004)   Laser  peripheral  iridotomy  

26.52A  Either  laser  peripheral  iridotomy  or  argon  laser  trabeculoplasty  or  selective  laser  trabeculoplasty  

End-­‐Stage  Glaucoma  (Laser  Treatment)   26.98B(2005-­‐2009)   Diode  laser  cyclophotocoagulation  

(ciliary  body  ablation)    Glaucoma  Surgery   26.2A  (1995-­‐2009)   Major  glaucoma  operation  

(trabeculectomy,  EMS  shunt)  

26.2B  (1995-­‐2009)   Ahmed  shunt  or  Baerveldt  shunt,  with  scleral  patch  graft  

26.25A  (1998-­‐2009)   Repeat  trabeculectomy  

 

Source:  Alberta  Diabetes  Atlas  2011  

 

 

 

 

 

   

85  

Table S3.2. Distribution of eye care services by household income in 2009

Household  income   Lowest  (%)   2nd  (%)   3rd  (%)   4th  (%)   Highest  (%)   Total  (%)  

SEX                Male   24.9   26.6   29.6   28.9   28.9   27.8  Female   27.3   31.0   31.6   30.1   30.5   30.1  

Age                Younger   21.9   22.3   24.5   24.8   24.0   23.5  Older   32.2   34.5   39.5   35.4   37.7   35.9  

Place  of  residence                Urban   27.8   33.2   28.3   30.5   30.0   30.0  Rural   21.2   26.2   18.7   30.0   25.2   24.3  

Aboriginal  status                Yes   18.4   16.6   22.7   19.7   21.4   19.8  No   26.9   28.4   31.1   30.2   29.6   29.2  

Duration  of  DM              Shorter   23.1   23.7   26.9   26.2   25.9   25.1  Longer   34.2   38.6   39.8   37.2   40.3   38.0  

Total   26.3   27.7   30.8   29.6   29.9   28.9  

           

                                                       

       

   

86  

Table S3.3. Distribution of eye care services by material deprivation index in 2009  

Material  deprivation   Worst  (%)   2nd  (%)   3rd  (%)   4th  (%)     Least  (%)   Total  (%)  

SEX                Male   25.0   27.4   27.2   26.9   32.6   27.8  Female   28.2   28.5   30.0   29.7   34.3   30.1  

Age                Younger   22.9   22.2   22.9   23.0   26.9   23.6  Older   31.6   35.1   35.5   36.3   40.6   35.8  

Place  of  residence                Urban   26.8   29.8   30.1   30.1   33.3   30.0  Rural   24.0   25.3   21.3   24.3   25.5   24.1  

Aboriginal  status                Yes   19.2   23.8   19.2   19.9   17.5   19.9  No   26.7   28.8   29.0   28.2   33.9   29.3  

Duration  of  DM                Shorter   22.7   24.8   24.8   24.7   28.9   25.2  Longer   35.0   36.1   37.5   38.2   43.5   38.1  

Total   26.0   28.6   28.3   28.4   33.2   28.9  

                                                         

 

   

87  

Table S3.4. Distribution of eye care services by social deprivation index in 2009  

Social  deprivation   Worst  (%)   2nd  (%)   3rd  (%)   4th  (%)     Least  (%)   Total  (%)  

SEX                Male   29.0   31.0   25.4   26.2   27.1   27.7  Female   32.1   34.8   26.9   27.7   29.3   30.1  

Age                Younger   24.3   23.8   23.7   22.6   23.1   23.5  Older   36.8   42.5   31.4   31.9   36.0   35.7  

Place  of  residence                Urban   30.9   30.1   32.5   27.3   29.0   30.0  Rural   26.2   21.3   19.9   25.3   28.3   24.2  

Aboriginal  status                Yes   17.9   20.1   19.6   21.0   21.5   20.0  No   30.8   33.6   27.4   25.6   28.8   29.3  

Duration  of  DM                Shorter   26.8   28.7   23.2   22.9   24.2   25.1  Longer   39.1   43.1   35.2   34.8   38.0   38.1  

Total   30.5   32.7   27.1   25.4   28.4   28.8  

 88  

Chapter 4. Factors Associated with Eye Screening Services and Visual Impairment in Canadians Living with Type 2 Diabetes*

Abstract

Diabetic retinopathy (DR) is the major cause of visual impairment and blindness in

diabetic patients. DR is known to be preventable and manageable, so regular eye

screening for early detection and timely treatment is recommended. Underutilization of

these services has been reported in Canada, although the impact of socioeconomic status

(SES) on utilization and visual impairment is not well documented. We used data from

the Survey on Living with Chronic disease in Canada (SLCDC)- Diabetes Component

2011, with respondents who self-reported having type 2 diabetes. Factors associated with

the use of eye screening and visual impairment (i.e., DR, cataracts, glaucoma) were

assessed using separate logistic regression models, weighted for the SLCDC sampling

strategy, to represent that Canadian population. Respondents to the SLCDC-DM with

type 2 diabetes (N=2,323) were 41.5% female, average age 63.4 (SD=0.4) years, with 44%

having more than 10 years duration of diabetes. Amongst all patients with type 2 diabetes,

factors associated with increased eye screening were discussion of diabetic complications

with health professionals (OR=2.02; 95%CI: 1.28- 3.19), having private insurance

(OR=3.23; 95%CI: 2.21- 4.73), and more than 10 years duration of diabetes (OR=1.53;

95%CI: 1.04 -2.25). Amongst patients with type 2 diabetes who reported having no visual

impairment, patients with lower income were 40% (OR=0.60; 95%CI: 0.37-0.98) less

likely to have eye screening compared to those with higher income. Amongst all patients

with type 2 diabetes, visual impairment was more likely in female (OR=1.53; 95%CI:

1.12-2.09), older patients (OR=18.12; 95%CI: 6.63-49.51), those with poor self-rated

                                                                                                                         *The research and analysis are based on data from Statistics Canada and the opinions expressed do not represent the views of Statistics Canada.  

08  Fall  

 89  

health (OR= 3.10; 95%CI: 1.62-5.96), and with lower income (OR=0.60; 95%CI: 0.37 -

0.98). Our study found that patient’s experience in discussing with health professionals

and having private health insurance were main factors associated with the use of eye

screening services, while age, sex, duration of diabetes, and self-rated health were related

to visual impairment among type 2 diabetic patients. These factors need to be considered

when the relevant health care services are provided to patients with type 2 diabetes.

 90  

4.1. Introduction

Diabetes retinopathy (DR) is the major cause of visual impairment and blindness,

eventually affecting almost all patients with type 1 diabetes and approximately 60~80%

of patients with type 2 diabetes (1-3). DR prevalence could be as high as 40% among

Canadian patients with diabetes (4) and it is anticipated that the prevalence of DR will

continuously increase as the number of diabetic patient escalates (5). DR is known to be

preventable and manageable if diagnosed at an early stage (6-8). Regular eye screening

for early detection and timely treatment decreases the risk of visual impairment and

blindness caused by DR (9).

The estimated cost of DR screening is much lower than the estimated cost of social

support and services related to visual impairment and blindness (2, 10). Thus, it is

economically beneficial to provide regular DR screening to patients with diabetes (8);

annual DR screening by an experienced health professional is strongly recommended for

all patients with diabetes (11). Despite the clinical efficiency and economic benefits of

regular screening, underutilization of DR screening services has been observed (2, 12, 13).

In Alberta, only 58% of people with diabetes received an eye exam by an

ophthalmologist within three years of diabetes onset (8). Similarly, in Ontario, only 19%

of diabetic patients have undertaken follow-up eye screening after an initial eye

examination (14).

Various barriers to receiving DR screening services have been identified, but lower

socioeconomic status (SES) is thought to be one of the major factors associated with the

failure to use eye screening services (15-19). Moss et al. found that higher education,

 91  

higher income, and health insurance for eye examination were predictors of eye

examinations for individuals in the US (20). Under the universal health care system in the

U.K., the rate of DR screening was significantly lower in patients residing in deprived

areas (16, 21, 22). Studies in Australia, Spain, and Hong Kong consistently found that

those with lower income were at higher risk of underutilization (18).

As there is a strong socioeconomic gradient in ill health (23), and eye screening is less

common in lower SES, it would follow that low SES is alsoa key predictor of increased

rates of diabetic eye complications (12, 21, 24-29). In a review of the literature, Brown et

al. found that lower SES, as measured by individual or household income, education,

employment, occupation, or living in an underprivileged area, was associated with

increased risk of microvascular disease, including diabetic eye complications (30). In a

community-based study conducted in the U.K., lower education was found to be a

predictor of diabetic eye complication, including DR (31).

There is little empirical evidence illustrating association between socioeconomic factors

and visual impairment as well as relationship between these factors and use of eye

screening services among Canadians living with diabetes. Therefore, the purpose of our

study was to examine whether socioeconomic factors are associated with visual

impairment and use of eye screening services in the Canadian diabetic population, using a

nationally representative survey data.

4.2. Methods

 92  

Data Source

Our study used data from the Survey on Living with Chronic Disease in Canada –

Diabetes Component 2011 (SLCDC-DM). The SLCDC is a cross-sectional survey

sponsored by the Public Health Agency of Canada regarding the experiences of

Canadians living with chronic health conditions (32). Individuals 20 years of age or older

who self-reported diabetes on the 2010 Canadian Community Health Survey (CCHS)

were invited to participate in the 2011 SLCDC-DM survey (32). Among the 3,590 CCHS

respondents contacted, 2,933 ultimately participated in the SLCDC (32). The SLCDC-

DM excluded full-time members of the Canadian Forces and residents of First Nations

Reserves, Crown lands, institutions, and the three territories (32). We included only

respondents who self-reported type 2 diabetes, based on the Ng, Dasgupta, and Johnson

(NDJ) classification algorithm (33). Respondents with any missing values were excluded

from our analysis.

Eye Screening Services

In the SLCDC survey, individuals were asked, “Have you ever had an eye exam where

the pupils of your eyes were dilated?” and “When was the last time (you had your pupils

dilated)?” Based on these two questions, we classified respondents into two groups: (1)

patients who have had an eye exam within less than 2 years, as a regular eye screening

group and (2) patients who have not had an eye exam or who had an eye exam 2 or more

years ago, as a non-regular eye screening group. We ran two different models for eye

screening services, first including all respondents with type 2 diabetes, and second

including only those reporting no visual impairment, to assess factors associated with

what could be considered preventive eye screening services.

 93  

Visual Impairment

SLCDC-DM participants were asked whether they have ever had eye complications

diagnosed by health professionals, including DR, cataracts, or glaucoma. We classified

patients who have been diagnosed with one of these eye complications as the “visual

impairment group” and those who have never had a diagnosed eye complication as the

“non-visual impairment group”.

Predictive Factors

Independent predictor variables were selected based on Andersen’s Health Behaviour

Model (HBM) and determinants of eye health from previous research (34). According to

Andersen’s HBM, the determinants of health care services can be conceptualized into

three categories: (1) predisposing characteristics, (2) enabling resources, and (3) need-

based factors (34-36). Predisposing characteristics are generally demographic factors

associated with the use of health care services such as sex, age, educational attainment,

and other socio-demographic factors (34). Enabling resources refer to resources that are

available to individuals for the use of health care services, such as insurance, availability

of physicians, and income (34). Need factors are indicators of actual health status that are

often associated with the use of health care services (36).

As predisposing factors we included age as categories (i.e., 20–49, 50–64, 65–79, 80 or

older), sex, and marital status (i.e., married or living with partner and single or living

alone), and educational attainment (i.e., less than secondary school, completed secondary

school, other postsecondary school, and college or university level of postsecondary

 94  

school). Respondents’ demographic information was included as: a) non-immigrant/non-

Aboriginal, b) immigrant, and c) Aboriginal. An individual’s spoken language at home

was also included.

As enabling resources we included self-reported household income (i.e.,< $15,000;

$15,000–29,999; $30,000–59,999; $60,000–79,999; and $80,000 or more). Private

insurance for eye care, and patients’ experience in discussing diabetic complications with

the family doctor were included in the model for eye screening services. We also

included place of residence in urban or rural dwelling, and living in one of the following

regions – Atlantic, Quebec, Ontario, Prairies, and British Columbia, as coverage and

service delivery models are likely to differ by region.

For need-based factor, we included duration of diabetes, dichotomized at more or less

than 10 years, and self-rated health classified as 4 categories: excellent/very good, good,

fair, poor. Visual impairment was included as a need-based factor covariate in the

analytic model for eye screening among all type 2 diabetic patients, but was excluded in

the second model for eye screening where we excluded all type 2 diabetic patients with

visual impairment. The re-categorization of all the variables was performed considering

Statistics Canada’s requirement of a minimum sample size in each dummy variable

category.

All descriptive and logistic regression analyses were conducted using STATA 12 for Mac.

Results are presented in terms of odds ratios (ORs) with 95% confidence intervals (CI)

 95  

and p-values. Differences were considered statistically significant at p<0.05. The survey

weights provided by Statistics Canada and bootstrap method were used for all analyses,

to account for the complex survey design and sample selections, adjustments for non

response, seasonal effects, and post-stratification (32). Weighted data are therefore

representative of the survey populations and are required by Statistics Canada for

reporting when producing population estimates (32).

4.3. Results

Of the 2,933 respondents to the SLCDC, 2,323 (79.2%) were classified as having type 2

diabetes and included in these analyses. Using the sample weights, this represented an

estimated total population of 1,324,553 type 2 diabetic patients in Canada. The general

characteristics of estimated population are summarized in Table 4.1. Among the total

population, 72% reported receiving a dilated eye screening services within 2 years (Table

4.1). There were little differences in eye screening services across sex, age, marital status,

education or ethnic/language groups. Across different income groups, the highest income

group ($80,000 or more) reported the highest rate, 78% of eye screening services, but the

differences were not statistically significant. Patients with longer duration of diabetes

reported a higher rate of the eye screening services. Among four different self-rated

health groups, the rate of eye screening services was higher in those reporting poor health.

In addition, individuals living in Quebec showed the lowest rate of a dilated eye

examination while those from Prairie provinces showed the highest rate of the

examination (58% vs. 79%; p<0.001). Patients who had discussed diabetic complications

with family physicians and who had private insurance for eye care showed higher rates of

eye screening services compared to their counterparts. Approximately 84% of patients

 96  

who had visual impairment had received eye screening services, while only 66% of

patients without visual impairment reported having eye screening services in the past 2

years.

One-third (33%) of type 2 diabetic patients in our study reported having visual

impairment including DR, cataract or glaucoma, as diagnosed by a health professional

(Table 4.1). Female patients reported a significantly higher prevalence of visual

impairment (40%) than men (28%). The prevalence of visual impairment increased

substantially with age, and was higher in respondents who were single or living alone.

Educational attainment and income were inversely related to the prevalence of visual

impairment. There was no difference in the prevalence of visual impairment across the

identified ethnic groups, but those respondents who spoke an official language at home

had higher prevalence than those speaking non-official languages (34% vs 20%; p=0.03).

Visual impairment was more common in patients with longer duration of diabetes (44%

vs24%; p<0.001), and in respondents who reported poor self-rated health (44%; p<0.001).

There were no differences in the prevalence of visual impairment between urban and

rural dwellers, or in the different regions of Canada. Discussion of diabetic complications

with health professionals and private insurance for eye care appointment were not related

to the prevalence of visual impairment.

Factors Related to Use of Eye Screening Services

 97  

In the first multivariable model including all type 2 diabetic patients, we found visual

impairment, duration of diabetes, having private insurance for eye care, and discussion of

diabetic complications with a health professional were the factors most strongly

associated with use of eye screening services (Table 4.2). Patients with visual impairment

were more than 2.5 (OR=2.60; 95%CI: 1.73- 3.91) times as likely, and patients living

with diabetes for more than 10 years were more than 1.5 (OR=1.53; 95%CI: 1.04- 2.25)

times as likely, to have received eye screening services within 2 years (Table 4.2).

Patients who have private health insurance for eye care were much more likely to have

eye screening services (OR= 3.23; 95%CI: 2.21-4.73). None of the socio-demographic,

income or area of residence factors were significantly associated with the use of eye care

services.

When we limited the analysis to type 2 diabetic patients without visual impairment, in

order to see the association between SES and preventive eye screening services, we found

similar results, in that duration of diabetes, discussing diabetic complications with health

professionals and having private insurance were strongly and significantly associated

with the use of eye screening services. However, there was a statistically significantly

greater likelihood for respondents with higher incomes to have preventive eye screening

services compared to those in the lowest income quintiles (Table 4.2). As a sensitivity

analysis, were classified income quintiles into dichotomous variable as lower (less than

$50,000) and higher (more than $50,000). In this analysis, those with lower income were

40% less likely (OR=0.60; 95%CI: 0.37-0.98) to have preventive eye screening services,

after adjusting for other demographic and socio-economic factors [Table S4.1].

 98  

Factors Related to Visual Impairment

We found that sex, age, duration of diabetes, self-rated health, income, and region of

Canada were significantly associated with visual impairment (Table 4.2). Female patients

were more likely to have visual impairment compared to male patients (OR = 1.53; 95%

CI: 1.12- 2.09). As might be expected, older patients were more likely to have visual

impairment, particularly patients over the age of65 years. Compared to respondents living

in Atlantic Canada, respondents in the other regions were more likely to report visual

impairments, with the highest rates in British Columbia (OR= 2.12; 95%CI: 1.13- 3.97)

and Ontario (OR= 1.72; 95%CI: 1.10- 2.69). Across income quintiles, individuals in the

higher income groups were 45 to 55% less likely to have visual impairment compared the

reference group, the lowest income quintiles (less than $15,000 per year) (Table 4.2).

4.4. Discussion

Regular eye screening is recommended for patients with type 2 diabetes in order to

promote early detection and timely treatment of diabetic eye complications (11). For this

reason, possible barriers for eye screening services and visual impairment should be

identified and ameliorated. Our findings, using a nationally representative survey dataset,

indicate that increased use of eye screening was associated with enabling factors such as

private insurance for eye screening services and household income. Importantly, however,

independent of these enabling socioeconomic factors, we found that discussing diabetic

complications with health professionals was associated with a greater likelihood of

receiving eye screening services. We also found that income was associated with the

prevalence of visual impairment amongst patients with type 2 diabetes, and that the

 99  

prevalence of visual impairment varied in regions across the country, with the lowest

rates in Atlantic Canada.

It has been well documented that people with higher levels of education or income are

more likely to use preventive or specialist care services compared to individuals with

lower levels of education or income (31, 37, 38). Our findings agree with accumulated

evidence related to the use of specialist care services that higher income is a predictor of

the use of preventive eye screening services. One possible explanation is the relationship

between income and a referral from a primary care physician, suggesting that individuals

with higher income tend to be well aware of the need to use preventive services, so they

are more likely to ask for a referral for these services (39, 40). Another explanation is that

an individual’s socioeconomic position has been shown to influence physicians’

perception of patients’ social class and intelligence (30). These perceptions may be

related to a physician’s preference to provide adequate counselling or referral services,

which would result in the tendency of patients of lower income not to use the

recommended eye screening services. There is consistent evidence identifyingincome as

an important factor in receiving a referral to specialist services (41). These findings

support the relationship between income and regular eye screening services in the

diabetic population as well.

The association between private health insurance and eye screening services among

patients with diabetes has been addressed in the literature. Several U.S. studies have

previously identified private health insurance as one of the barriers to access eye

screening services (20, 30, 42). Diabetic patients without insurance covering eye

 100  

screening services are less likely to receive the recommended screening services that are

commonly expectedin a private health care setting (42). However, even in Canada, our

finding demonstrates that private insurance for eye care services is a predictor of regular

eye screening services among patients with diabetes. This finding could be explained by

the fact that eye screening services by an optometrist are sometimes not covered by

provincial health care and require private insurance. Having private insurance for eye

care appointments not covered by provincial insurance facilitates more frequent use of

health care services. In Canada, as private health insurance is usually provided through

the employer or more frequently found in higher paying positions, failing to cover eye

screening through the publicly funded health care system could create inequity in health

care for unemployed and precariously employed individuals (43).

Our study found that patients’ experience in discussion of diabetic complications with

health professionals is also an important indicator of receiving recommended eye

screening services. This finding emphasizes the importance of the role of health

professionals in diabetes management, in particular for eye screening services. One

qualitative study found that detailed information about eye screening from primary care

providers is an important factor in patient’s receiving eye screening services (44, 45). A

comprehensive primary care model has been suggested in order to enhance the

responsibility of primary care providers and develop an effective collaboration between

primary care providers and eye care services providers (46). Empirical evidence in the

U.K. found that one such model, the Diabetes Managed Clinical Network, has made

positive changes in collaborating closely with optometry and ophthalmology (45).

 101  

Clinical risk factors such as sex, age and duration of diabetes are associated with visual

impairment. This finding is in line with accumulative epidemiological studies for visual

malfunction in diabetic patients (47-52). Well-established clinical evidence indicates that

the risk of eye disease such as DR, glaucoma, and cataract increases in both prevalence

and severity in older age groups (47-49). Our study found a higher likelihood of visual

impairment in females compared to males. This association remained significant after

adjusting for various socioeconomic factors. Previous evidence on the association

between sex and visual impairment among diabetic patients is controversial. The

estimated prevalence of DR and vision-threatening DR was similar in male and female

groups in a study estimating global prevalence of DR in diabetic population (50). Another

study reported that males represent an independent risk factor for severe DR as well as

early DR in patients with type 2 diabetes (51). In contrast, and consistent with our

findings, data from the US National Society to Prevent Blindness (NSPB) indicated that

the prevalence rate for diabetes-related visual impairment and legal blindness tended to

be higher in females than in males (52).

Lower income is generally associated with visual impairment, suggesting that diabetic

patients with lower income might have more visual impairment than diabetic patients

with higher income (53). There is considerable evidence that socioeconomic status may

determine the risk of diabetes-related complications such as retinopathy that ultimately

have an impact on the quality of life (15, 21). There is evidence that diabetic patients of

lower socioeconomic status, as measured by individual or household income, have

increased risk of eye complications (30). Our findings support previous researchthat

income is a predictor of visual impairment among diabetic patients (30). The causality

between income and eye complication is still unclear but it can be argued that patients

 102  

with visual impairment may impede participation in the labour force market, resulting in

lower income. On the other hand, previous studies have suggested that health is generally

influenced by income and is shaped over time by the socioeconomic status imposed on

individuals at different stages of life (23, 54).

The higher prevalence of visual impairment in Ontario and British Columbia compared to

Atlantic Canada is difficult to explain. While we controlled for age, sex, income and

duration of diabetes, other factors not adequately controlled in our model might account

for some of the difference, also called residual confounding. It did not appear that use of

eye screening services differed across the same regions. Further study is required to

understand the observed higher prevalence of visual impairment amongst diabetic

individuals living in Ontario and British Columbia compared to other Canadian provinces.

Limitation

Our study has a number of limitations. First, due to the snapshot nature of the cross-

sectional design, causal inferences cannot be inferred between various factors, including

SES, and eye screening services, or visual impairment. The nature of the data collection

also precludes the exclusion of reverse causality in our study findings. Second, the

SLCDC- DM 2011 is a self-reported survey data and therefore prone to measurement

error and biases, such as recall and social desirability bias. The latter is quite likely, given

that the self-reported rates of regular eye screening services are higher than previous

Canadian studies (8, 14). Another potential limitation is that sampling bias may limit the

generalizability of findings; for example, residents of three territories and of First Nations

reserves were not included in the SLCDC- DM 2011. Finally, the self-reported nature of

 103  

the data does not provide clinical parameters, such as glycemic control, blood pressure,

which would reflect the level of risk for diabetic eye disease.

4.5. Conclusion

Findings from our study highlight the importance of socioeconomic factors in

determining the use of eye screening services and the risk of visual impairment,

regardless of other factors that may confound or mediate these associations. Diabetic

retinopathy is a treatable and preventable diabetic complication. Regular screening for

diabetic eye disease and timely treatment are clinically effective and economically

beneficial. In order to provide an eye screening and appropriate treatment to type 2

diabetic patients, the factors associated with use of eye screening services and visual

impairment we identified need to consider when ophthalmology or optometry services are

provided to type 2 diabetic patients.

 

 

 

 

 104  

Tables  

Tab

le 4

.1. C

hara

cter

istic

s of

type

2 d

iabe

tic p

atie

nts

by e

ye sc

reen

ing

serv

ices

Cha

ract

eris

tics

C

ateg

ory

Tot

al e

stim

ate

popu

latio

n E

ye s

cree

ning

P-

vaul

eb V

isua

l Im

pair

men

ta P-

valu

eb N

%

%

- Y

es

%-Y

es

Tot

al

Popu

latio

n

1324

553

100.

0 72

.0

32

.7

Sex

Mal

e 77

4810

58

.5

70.7

0.

38

27.8

0 <0

.001

Fe

mal

e 54

9742

41

.5

73.7

39

.70

Age

20-4

9 18

8680

14

.2

70.2

0.51

10.7

0

<0.0

01

50-6

4 53

8478

40

.7

69.9

21

.20

65-7

9 49

3328

37

.2

75.2

45

.90

80+

1040

67

7.9

70.5

69

.90

Mar

ital s

tatu

s M

arri

ed/p

artn

ered

92

7550

70

.0

72.5

0.

55

27.6

0 <0

.001

Si

ngle

/livi

ng a

lone

39

7003

30

.0

70.6

44

.60

Edu

catio

n

Les

s th

an S

econ

dary

19

8249

15

.0

71.8

0.72

47.1

0

<0.0

01

Seco

ndar

y 16

6632

12

.6

68.1

27

.80

Oth

er p

ost-

seco

ndar

y 86

964

6.6

75.8

31

.50

Post

-sec

onda

ry

8727

08

65.9

72

.3

30.5

0

Dem

ogra

phic

Non

-im

mig

rant

/non

-A

bori

gina

l 93

8678

70

.9

71.3

0.

80

34.2

0 0.

29

Imm

igra

nt

3387

60

25.6

73

.6

28.4

0 A

bori

gina

l 47

115

3.6

27.0

34

.80

Off

icia

l la

ngua

ge

at h

ome

Yes

12

1223

0 91

.5

72.1

0.

87

33.9

0 0.

03

No

1123

23

8.5

70.9

19

.60

Vis

ual

impa

irm

ent

Yes

43

3246

32

.7

83.7

<0

.001

N

/A

No

8913

07

67.3

66

.3

!

 105  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Dur

atio

n of

D

M

Less

than

10y

rs

7358

95

55.6

66

.8

<0.0

01

23.5

0 <0

.001

M

ore

than

10y

rs

5886

58

44.4

78

.4

44.2

0

Self-

rate

d H

ealth

Exce

llent

& v

ery

good

32

6026

24

.6

73.2

0.

09

23.4

0

<0.0

01

Goo

d 55

5678

42

.0

69.3

30

.60

Fair

2869

32

21.7

69

.9

41.0

0 Po

or

1559

17

11.8

82

.6

44.3

0

Inco

me

Less

than

$15

,000

96

125

7.3

71.0

0.34

54.0

0

<0.0

01

$15,

000~

29,9

99

2496

31

18.8

70

.9

45.1

0 $3

0,00

0~49

,999

33

4876

25

.3

69.8

31

.60

$50,

000~

79,9

99

3410

34

25.7

69

.8

27.1

0 $8

0,00

0 or

mor

e 30

2887

22

.9

78.0

23

.30

Res

iden

ce

Urb

an

1018

659

76.9

73

.1

0.22

33

.70

0.21

R

ural

30

5894

23

.1

68.3

29

.40

Reg

ion

Atla

ntic

11

7450

8.

9 72

.2

<0.0

01

27.9

0

0.56

Q

uebe

c 27

2315

20

.6

57.5

31

.50

Ont

ario

60

1503

45

.4

75.6

33

.40

Prai

rie

1963

30

14.8

78

.9

31.5

0 B

C

1369

55

10.3

74

.5

38.0

0

D

iscu

ssio

n w

ith

Hea

lth

prof

essi

onal

s

Yes

10

7509

4 81

.2

75.8

<0.0

01

32.7

0

0.99

No

2494

59

18.8

55

.5

32.7

0

Priv

ate

Insu

ranc

e Y

es

8947

91

67.6

80

.7

<0.0

01

34.0

0 0.

23

No

4297

62

32.4

53

.8

30.0

0 ! a  visual  im

pairment  included  self-­‐reported  diabetic  retinopathy,  cataracts  or  glaucom

a  

b  univariate  com

parison  with  chi-­‐square  test  

 

 106  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable(

Eye$screening$services$

Preventive$eye$screening$$

Visual$Im

pairment$

OR(

95%(CI((

P1value(

OR

95%(CI((

P-va

lue

OR

95%(CI(

P-va

lue

Predisposing((

Characteristic(

Sex$

$$!

$$$

$$$

$$$

$$Female$

1.22

0.83$:1.78$

0.31

1.28$

0.80$:$2.05$

0.31$

1.53$

1.12$:$2.09$

0.01*$

Age$

$$

$$$

$$

$$$

$50:64$

0.88

0.43$:$1.79$

0.22

0.85$

0.38$:$1.90$

0.68$

1.79$

0.71$:$4.47$

0.22$

65:79$

0.98

0.49$:$1.96$

0.14

1.15$

0.51$:$2.63$

0.73$

6.31$

2.47$:$16.11$

0.00*$

80:$

0.86

0.38$:$1.95$

0.21

1.08$

0.37$:$3.20$

0.88$

18.12$

6.63$:$49.51$

0.00*$

Marital$status$

$$

$$$

$$

$$$

$Single$

0.84

0.57$:$1.24$

0.38

0.82$

0.51$:$1.33$

0.42$

1.42$

0.99$:$2.03$

0.05$

Education$

$$

$$$

$$

$$$

$Secondary$

0.82

0.46$:$1.46$

0.51

0.61$

0.29$:$1.26$

0.18$

0.61$

0.36$:$1.02$

0.06$

Other$post:sec$

1.16

0.55$:2.47$

0.70

0.94$

0.41$:$2.16$

0.89$

0.92$

0.48$:$1.76$

0.80$

Post:sec$

1.01

0.63$:$1.62$

0.97

0.77$

0.43$:$1.36$

0.36$

1.26$

0.86$:$1.85$

0.24$

Demographic$

$$

$$$

$$

$$$

$Immigrant$

1.00

0.56$:$1.78$

1.00

1.10$

0.53$:$2.26$

0.80$

0.89$

0.54$:$1.46$

0.64$

Aboriginal$

0.73

0.34$:$1.58$

0.42

0.78$

0.27$:$2.23$

0.64$

0.99$

0.45$:$2.19$

0.98$

Official$Language$$

$$

$$$

$$$

$$

No$

1.14

0.47$:$2.77$

0.78

1.00$

0.33$:$3.03$

0.99$

0.45$

0.16$:$1.28$

0.13$

Need(

factors(

Visual$

impairment$

2.60

1.73$:$3.91$

0.00

* $$

$$$$

$$$$

$$

Yes$

$$

$$$

$$

$$$

$Duration$of$DM$

$$

$$$

$$

$$$

$>$10yrs$

1.53

1.04$:$2.25$

0.03

* 1.53$

0.93$:$2.50$

0.09$

2.14$

1.58$:$2.89$

0.00*$

Self<rated$health$

$$

$$$

$$$

$$

Good$

0.75

0.48$:$1.16$

0.20

0.74$

0.44$:$1.26$

0.27$

1.47$

0.98$:$2.21$

0.06$

Fair$

0.69

0.41$:$1.16$

0.16

0.65$

0.35$:$1.23$

0.18$

2.03$

1.27$:$3.24$

0.00*$

Poor$

1.40

0.64$:$3.08$

0.40

1.76$

0.60$:$5.13$

0.30$

3.10$

1.62$:$5.96$

0.00*$

$Tab

le 4

.2. M

ultiv

aria

te lo

gist

ic re

gres

sion

for e

ye s

cree

ning

ser

vice

s, p

reve

ntiv

e ey

e sc

reen

ing

serv

ices

and

vis

ual i

mpa

irmen

t

 

 107  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Enabling(

facto

rs(

Income'

!!

!!!

!!!!

!!!!

!!"29

,999!

1.27

0.64!"!2.5

1!0.

50

2.22!

1.02!"!4.8

5!0.04*!

0.66!

0.40!"!1.1

1!0.12!

"49,999!

1.12

0.55!"!2.2

7!0.

76

1.47!

0.68!"!3.1

8!0.33!

0.50!

0.28!"!0.8

7!0.01*!

"79,999!

1.13

0.51!"!2.4

7!0.

77

1.51!

0.64!"!3.5

6!0.34!

0.45!

0.24!"!0.8

3!0.01*!

80,000!m

ore!

1.51

0.66!"!3.4

4!0.

33

2.19!

0.87!"!5.5

2!0.10!

0.56!

0.27!"!1.1

6!0.12!

Urban/rural'

!!

!!!

!!

!!!

!Urban

!1.

07

0.72!"!1.5

9!0.

73

1.07!

0.67!"!1.7

2!0.77!

1.05!

0.70!"!1.5

7!0.81!

Region'

!!

!!!

!!

!!!

!Quebe

c!0.

86

0.47!"!1.5

5!0.

61

0.76!

0.37!"!1.5

5!0.45!

1.42!

0.88!"!2.3

0!0.15!

Ontario!

1.03

0.64!"!1.6

6!0.

91

1.09!

0.63!"!1.8

9!0.75!

1.72!

1.10!"!2.6

9!0.02*!

Prairie

!!1.

32

0.81!"!2.1

6!0.

26

1.50!

0.82!"!2.7

2!0.18!

1.36!

0.82!"!2.2

7!0.24!

BC!

1.07

0.58!"!1.9

7!0.

83

1.26!

0.60!"!2.6

4!0.55!

2.12!

1.13!"!3.9

7!0.02*!

Discu

ssion'with'HP'

!!

!!!

!!!

!!

Yes!

2.02

1.28!"!3.1

9!0.

00*

2.54!

1.42!"!4.5

3!0.00*!

1.39!

0.96!"!2.0

2!0.08!

Private'Insurance'

!!

!!!

!!!

!!

Yes!

3.23

2.21!"!4.7

3!0.

00*

3.71!

2.33!"!5.9

1!0.00*!

1.12!

0.81!"!1.5

6!0.48!

!

 108  

References

1. Fong DS, Aiello LP, Ferris FL, Klein R. Diabetic Retinopathy. Diabetes care. 2004;27(10):2540-53. 2. Jones S, Edwards RT. Diabetic retinopathy screening: a systematic review of the economic evidence. Diabet Med. 2010;27(3):249-56. 3. Buhrmann RH, W. Beardmore, J. Baker,G. Lowcock, B. Pan, I. Bovell, A. Section 2. The Epidemilogoy of Blindness and Visual Imparment in Canada : Diabetic Retinopathy. Foundations for a Canadian Vision Health Strategy : Towards Preventing Avoidable Blindness and Promoting Vision Health: The National Coalition for Vision Health; 2007. p. 12-26. 4. Determining the information needs for prevention of vision loss strategies in Canada. Public Health Task Force for Vision and Ophthalmology. Can J Ophthalmol. 2000;35(5):255-7. 5. Aiello LCJB, SE. Diabetic eye disease. Endocrinol Metab Clin North Am. 1996;25(2):272-90. 6. Hooper P, Boucher MC, Cruess A, Dawson KG, Delpero W, Greve M, et al. Canadian Ophthalmological Society evidence-based clinical practice guidelines for the management of diabetic retinopathy. Can J Ophthalmol. 2012;47(2 Suppl):S1-30. 7. van Eijk KN, Blom JW, Gussekloo J, Polak BC, Groeneveld Y. Diabetic retinopathy screening in patients with diabetes mellitus in primary care: Incentives and barriers to screening attendance. Diabetes Res Clin Pract. 2011. 8. Rudnisky CJ, Tennat MTS, Johnson JA, & Balko SU. Chapter 9 : Diabetes and Eye Disease in Alberta. In: Johnson JA,editor Alberta Diabetes Atlas 2011. Edmonton: Institute of Health Economics. 2011. p. 153-74. 9. World Health Organization. Prevention of blindness from diabetes melitus. Geneva, Switzerland:World Health Orgnization.2005. 10. Zhang X, Norris SL, Saadine J, Chowdhury FM, Horsley T, Kanjilal S, et al. Effectiveness of interventions to promote screening for diabetic retinopathy. Am J Prev Med. 2007;33(4):318-35. 11. Bond SR AA, Altomare F., Stockl F. Canadian Diabetes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada: Retinopathy. Can J Diabetes. 2013;37(suppl1):S137-S41. 12. Millett C, Dodhia H. Diabetes retinopathy screening: audit of equity in participation and selected outcomes in South East London. J Med Screen. 2006;13(3):152-5. 13. Garg S, Davis RM. Diabetic Retinopathy Screening Update. Clinical Diabetes. 2009;27(4):140-5.

 109  

14. Burhrmann R, Assaad D, Hux JE, Tang M, & Sykora K. Chapter 10: Diabetes and the Eye. Diabetes in Ontario: An ICES Practice Atlas: Institute for Clinical Evaluative Sciences; 2003. p. 10.193- 10.208. 15. Scanlon PH, Carter SC, Foy C, Husband RF, Abbas J, Bachmann MO. Diabetic retinopathy and socioeconomic deprivation in Gloucestershire. J Med Screen. 2008;15(3):118-21. 16. Nsiah-Kumi P, Ortmeier SR, Brown AE. Disparities in diabetic retinopathy screening and disease for racial and ethnic minority populations--a literature review. J Natl Med Assoc. 2009;101(5):430-7. 17. He XZ. Diabetes preventive services and policy implications in the U.S. Diabetes care. 2011;34(1):8-13. 18. Bragge P, Gruen RL, Chau M, Forbes A, Taylor HR. Screening for presence or absence of diabetic retinopathy: a meta-analysis. Arch Ophthalmol. 2011;129(4):435- 44. 19. Browning DJ. Chapter 14. Screening for Diabetic Retinopathy. Diabetic Retinopathy. New York: Springer; 2010. p. 369-85. 20. Moss SE, Klein R, Klein BEK. Factors Associated With Having Eye Examinations in Persons With Diabetes. Arch Fam Med. 1995;4(6):529-34. 21. Kliner M, Fell G, Gibbons C, Dhothar M, Mookhtiar M, Cassels-Brown A. Diabetic retinopathy equity profile in a multi-ethnic, deprived population in Northern England. Eye (Lond). 2012. 22. Hippisley-Cox J, O'Hanlon S, Coupland C. Association of deprivation, ethnicity, and sex with quality indicators for diabetes: population based survey of 53,000 patients in primary care. BMJ. 2004;329(7477):1267-9. 23. Marmot M. The Marmot review final report: fair society, healthy lives. 2010. University College London. 24. Chaturvedi N, Stephenson J, Fuller J, Complications TEI. The Relationship Between Socioeconomic Status and Diabetes Control and Complications in the EURODIAB IDDM Complications Study. Diabetes care. 1996;19(5):423-30. 25. Fang R, Kmetic A, Millar J, Drasic L. Disparities in Chronic Disease Among Canada's Low-Income Populations. Prev Chronic Dis. 2009;6(4):A116. 26. Garcia AA, Benavides-Vaello S. Vulnerable populations with diabetes mellitus. Nurs Clin N Am. 2006;41(4):605-23, vii. 27. Gulliford MC, Dodhia H, Chamley M, McCormick K, Mohamed M, Naithani S, et al. Socio-economic and ethnic inequalities in diabetes retinal screening. Diabet Med. 2010;27(3):282-8. 28. Peek M, Cargill A, Huang E. Diabetes Health Disparities. Med Care Res Rev.

 110  

2007;64(5 suppl):101S-56S. 29. Rabi DM, Edwards AL, Southern DA, Svenson LW, Sargious PM, Norton P, et al. Association of socio-economic status with diabetes prevalence and utilization of diabetes care services. BMC health services research. 2006;6:124. 30. Brown AF, Ettner SL, Piette J, Weinberger M, Gregg E, Shapiro MF, et al. Socioeconomic Position and Health among Persons with Diabetes Mellitus: A Conceptual Framework and Review of the Literature. Epidemiol Rev. 2004;26(1):63- 77. 31. Bachmann MO, Eachus J, Hopper CD, Davey Smith G, Propper C, Pearson NJ, et al. Socio-economic inequalities in diabetes complications, control, attitudes and health service use: a cross-sectional study. Diabet Med. 2003;20(11):921-9. 32. Statistics Canada. Survey on Living with Chronic Diseases in Canada User guide. Ottawa; Canada: Statistics Canada, 2011. 33. Ng E, Dasqupta K, Johnson J. An algorithm to differentiate diabetic respondents in the Canadian Community Health Survey. Health Reports (Statistics Canada Catalogue 82-003). 2008;19(1). 34. Andersen RM. Revisiting the Behavioral Model and Access to Medical Care: Does It Matter? J Health Soc Behav. 1995;36(1):1-10. 35. Asada Y, Kephart G. Equity in health services use and intensity of use in Canada. BMC health services research. 2007;7(1):1-12. 36. Sibley L, Weiner J. An evaluation of access to health care services along the rural- urban continuum in Canada. BMC health services research. 2011;11(1):20. 37. Jimenez-Rubio D, Smith PC, Van Doorslaer E. Equity in health and health care in a decentralised context: evidence from Canada. Health economics. 2008;17(3):377-92. 38. Sara A. Does Equity in Healthcare Use Vary across Canadian Provinces? Healthcare Policy. 2008;3(4):83-99. 39. McGrail KM. Income-related inequities: Cross-sectional analyses of the use of medicare services in British Columbia in 1992 and 2002 Open Med.2008;2(4):E3-10. 40. Lorant V, Boland B, Humblet P, Deliege D. Equity in prevention and health care. J Epi Commu Health. 2002;56(7):510-6. 41. Dunlop S, Coyte PC, McIsaac W. Socio-economic status and the utilisation of physicians' services: results from the Canadian National Population Health Survey. Social Sci & Med. 2000;51(1):123-33. 42. Farley TF, Mandava N, Prall FR, Carsky C. Accuracy of Primary Care Clinicians in Screening for Diabetic Retinopathy Using Single-Image Retinal Photography. AFM. 2008;6(5):428-34.

 111  

43. Stabile M. Private insurance subsidies and public health care markets: evidence from Canada. Can J Economics. 2001;34(4):921-42. 44. Yeo ST, Edwards RT, Luzio SD, Charles JM, Thomas RL, Peters JM, et al. Diabetic retinopathy screening: perspectives of people with diabetes, screening intervals and costs of attending screening. Diabet Med. 2012;29(7):878-85. 45. Hayden C. The barriers and enablers that affect access to primary and secondary eye care services across England, Wales, Scotland and Northern Ireland.Royal National Insitute of Blind People. 2012. 46. Nam S, Chesla C, Stotts NA, Kroon L, Janson SL. Barriers to diabetes management: Patient and provider factors. Diabetes research and clinical practice. 2011;93(1):1-9. 47. Williams R, Airey M, Baxter H, Forrester J, Kennedy-Martin T, Girach A. Epidemiology of diabetic retinopathy and macular oedema: a systematic review. Eye. 2004;18(10):963-83. 48. Klein BEK. Overview of Epidemiologic Studies of Diabetic Retinopathy. Ophthalmic Epi. 2007;14(4):179-83. 49. O'Donnell K. Eye care in the UK : Epidemiology, intervention and ethnicity. Public Health Action Support Team, 2009 50. Yau JWY, Rogers SL, Kawasaki R, Lamoureux EL, Kowalski JW, Bek T, et al. Global Prevalence and Major Risk Factors of Diabetic Retinopathy. Diabetes care. 2012;35(3):556-64. 51. Kostev K, Rathmann W. Diabetic retinopathy at diagnosis of type 2 diabetes in the UK: a database analysis. Diabetologia. 2013;56(1):109-11. 52. Klein R, Klein BEK. Chapter 14. Visual disorders in diabetes. National Institute of Health, 1995. 53. Perruccio AV, Badley EM, Trope GE. A Canadian population-based study of vision problems: assessing the significance of socioeconomic status. Can J Ophthalmol. 2010;45(5):477-83. 54. Dinca-Panaitescu S. Dinca-Panaitescu M, Bryant T, Daiski I, Pilkington B, Raphael D. Diabetes prevalence and income: Results of the Canadian Community Health Serevy. Health Policy. 2011;99(2):116-23.

 112  

Supplementary Table

Table S4.1. Multivariate logistic regression for preventive eye screening services in type 2 diabetic patients

Variable   OR   95%  CI   P-­‐value  

Predisposing    Characteristic  

Sex          Female   1.28   0.80  -­‐  2.06   0.30  

Age        50-­‐64   0.75   0.34  -­‐  1.68   0.49  65-­‐79   1.14   0.50  -­‐  2.59   0.75  80-­‐   1.15   0.40  -­‐3.25   0.80  

Marital  status        Single   0.84   0.53  -­‐  1.26   0.34  

Education        Secondary   0.56   0.27  -­‐  1.16   0.12  

Other  post-­‐sec   0.82   0.35  -­‐  1.92   0.65  Post-­‐sec   0.65   0.36  -­‐  1.15   0.14  

Demographic        Immigrant   1.22   0.61  -­‐  2.45   0.57  Aboriginal   0.8   0.28  -­‐  2.30   0.68  

Official  Language  at  home      No   0.98   0.33  -­‐  2.89   0.99  

Need  factors  

Duration  of  DM              more  than  10yrs   1.52   0.93  -­‐  2.48   0.36  Self-­‐rated  health        

Good   0.79   0.47  -­‐  1.31   0.36  Fair   0.68   0.37  -­‐  1.26   0.23  Poor   2.03   0.66  -­‐  6.26   0.22  

Enabling    factors  

Income        Less  than  50,000   0.6   0.37  -­‐  0.98   0.04*  Urban/rural        

Urban   1.12   0.70  -­‐  1.80   0.62  Region        

Quebec   0.75   0.37  -­‐  1.53   0.43  Ontario   1.03   0.60  -­‐  1.75   0.92  Prairie     1.44   0.80  -­‐  2.60   0.22  BC   1.21   0.58  -­‐  2.51   0.62  

Discussion  with  HP        Yes   2.4   1.36  -­‐  4.24   0.00*  

Private  Insurance        Yes   3.46   2.20  -­‐  5.45   0.00*  

 

113  

Chapter 5. Income-related Disparities in Visual Impairment and Eye Screening Services in Type 2 Diabetic Patients in Canada*

Abstract

Diabetic retinopathy (DR) is the major cause of visual impairment and blindness among

working age individuals with diabetes. For early detection and timely treatment, regular

eye screening for DR is strongly recommended for all diabetic patients. Despite the

availability of eye screening services under Canada’s universal health care system,

underuse of routine eye screening services is observed. Among various factors, it is

acknowledged that income plays a crucial role in health and health care. The purpose of

this study was to measure income-related disparities in visual impairment and use of eye

screening services amongst Canadian living with type 2 diabetes. In addition, the study

aimed to identify contribution of various socio-demographic determinants to income-

related disparities. Our study used data from the Survey on Living with Chronic Disease

in Canada- Diabetes Component 2011 (SLCDC-DM). A total of 2,323 patients with type

2 diabetes were included in our study. To examine income-related disparities in visual

impairment and use of eye screening services, we used the relative concentration index

(RCI), horizontal inequity index (HI) and decomposition of the RCI. Our results suggest

that individuals with lower income tend to have more visual impairment compared to

those with higher income. The main contribution to the observed income disparity in

visual impairment came from age and marital status, suggesting the observed disparities

in visual impairment could be interpreted as an inequality because age and marital status

are more likely to be considered as unmodifiable factors. With respect to eye screening

services, patients with higher income were more likely to use more eye screening and

                                                                                                               *  The research and analysis are based on data from Statistics Canada and the opinions expressed do not represent the views of Statistics Canada.  

 

114  

preventive eye screening services. The main contributors of the observed disparities were

income, having private health insurance and patient’s experience in discussing diabetic

complications with health professionals. After health care need standardization, HIs still

indicate “pro-rich”, suggesting the observed disparities in eye screening services could be

the result of inequities. Our findings suggest that the identified determinants could be

considered when health and health care policies are developed in order to tackle

disparities in visual impairment and the use of eye screening services in diabetic

population.

 115  

5.1. Introduction

Diabetic retinopathy (DR) is the leading cause of visual impairment and blindness among

working age individuals with diabetes (1-3). DR is also associated with worse health

outcomes, which may ultimately impede the patient’s participation in economic activity

and decrease his or her quality of life (4, 5). Vision impairment caused by DR can be

preventable and is more manageable when the disease is diagnosed and treated at an early

stage; however, once vision loss develops, it cannot be recovered (6-8). Therefore,

regular eye screening by a health professional is strongly suggested for all diabetic

patients for early detection and timely treatment (2, 7).

DR is the primary cause of blindness in North America (9). It is responsible for about

12% of all new cases of blindness, affecting more than 8,000 individuals each year in the

U.S (10). In Canada, DR causes an estimated 600 new cases of blindness each year (11).

Considering the increasing number of diabetic patients in the Canadian population, the

prevalence of DR and blindness is expect to continuously increase (12). One Canadian

study has estimated that DR prevalence could be as high as 40% among diabetic patients

(10). It has also been estimated that nearly half a million Canadians currently have some

form of DR and approximately 100,000 Canadians have a more advanced form of the

disease defined as proliferative DR, diabetic macular edema or both (9). Glaucoma and

cataracts also occur earlier and more frequently, and progress more rapidly, in patients

with diabetes (13).

Cumulative evidence suggests that regular screening for DR is clinically effective and

economically beneficial (2, 14). This service is available under Canada’s universal health

care system, which aims to eliminate financial barriers to health care services. Despite the

availability of eye screening services within the publicly funded health care system,

 116  

underuse of routine eye screening services is reported (15, 16). In a study that estimated

the use of eye care services in five provinces and one territory, only 68% of diabetic

patients had received the recommended level of eye screening service (17). In Alberta,

only 55% of patients received an eye examination by an ophthalmologist within three

years of being diagnosed with diabetes, and the yearly number of eye examinations by

ophthalmologists has decreased over the last 15 years (2). In Ontario, only 19% of

diabetic patients have undertaken follow-up eye screening after their initial eye

examinations (18).

Among various factors, income has been identified as a key factor associated with both

visual impairment and use of eye screening services (19-22). In a literature review,

Brown et al. found that low socioeconomic status (SES), as measured by income, is

associated with increased risk of microvascular disease including ocular complications

(19). Moss et al. found diabetic patients in higher income groups were more likely to

receive eye screening services compared to those in lower income groups (20). Despite a

bourgeoning literature revealing income-related inequities in health and health care,

however, there is little evidence available regarding visual impairment and use of eye

screening services among type 2 diabetic patients.

The purpose of this study was to measure income-related disparities in visual impairment

and use of eye screening services amongst Canadians living with type 2 diabetes. Also we

aimed to identify the contribution of various socio-demographic determinants to income-

related disparities in visual impairment and use of eye screening services.

5.2. Methods

Data source

 117  

Our study used data from the Survey on Living with Chronic Disease in Canada—

Diabetes Component 2011 (SLCDC-DM). The SLCDC is a cross-sectional survey on the

experiences of Canadians living with chronic health conditions (23). Individuals 20 years

or older who self-reported diabetes diagnosed by a health professional in the 2010

Canadian Community Health Survey (CCHS) were invited to participate in the 2011

SLCDC-DM (23). The SLCDC-DM excludes individuals who are full-time members of

the Canadian Forces and residents of First Nations reserves, Crown islands, institutions,

and the three territories (23). 2,933 individuals (out of a total of 3,590 CCHS respondents

who were eligible) responded and participated in the 2011 SLCDC-DM survey. We only

included respondents who self-reported type 2 diabetes, based on the Ng, Dasgupta and

Johnson (NDJ) classification algorithm (24). Respondents with any missing value were

excluded from our analysis.

Outcome variables – visual impairment and eye screening services

In order to assess visual impairment, SLCDC-DM survey participants were asked the

following question: “Have you ever had any of the following conditions diagnosed by a

health professional: diabetic eye disease or diabetic retinopathy; partial or complete

blindness; cataracts; glaucoma?”. Patients who reported one of these eye complications

were classified as individuals with visual impairment and patients who did not report any

eye complication were classified as individuals with no visual impairment.

Participants were also asked to self-report (i) use of dilated eye screening services and (ii)

the last time a dilated eye screening service was used. Patients who had undergone dilated

eye screening within two years were classified as regular eye screening receivers whereas

patients who had not undergone dilated eye screening or had undergone dilated eye

screening two or more years before our study were defined as non-regular eye screening

 118  

receivers. We developed two different models for eye screening services, first including

all respondents with type 2 diabetes, and second including only those reporting no visual

impairment.

Socioeconomic factors

We used total household income as the measure of socio-economic status. Respondents

were asked to estimate their total household income in the past 12 months. This self-

reported income was categorised into 10 income decile groups based on the adjusted ratio

of individual’s total household income to the low income cut-off corresponding to

household and community size (25). For our decomposition models, socio-demographic

covariates for use of eye screening services and visual impairment were selected based on

Andersen’s Health Behaviour Model (HBM) and determinants of health care utilization

and health outcomes from previous studies (26). These variables included age group (i.e.,

20-49, 50-64, 65-79, 80 or older), sex, marital status (i.e., married or living with partner

and single or living alone), demographic status (i.e., non-immigrants, Aboriginals and

immigrants), and duration of diabetes. In addition, educational attainment (i.e., less than

secondary school, completed secondary school, other postsecondary school, and college

or university level of postsecondary school), place of residence (i.e., urban or rural) and

region of Canada, self-rated health (i.e., excellent/very good, good, fair, bad), whether

private insurance for eye care appointments was available, and experience discussing

diabetic complications with a health professional were included variables.

The Relative Concentration Index (RCI) and decomposition analysis

To examine income-related disparities in visual impairment and use of eye screening

services, we used the Relative Concentration Index (RCI). The RCI was proposed by

Kakwani and Wagstaff (27, 28) and is widely used as a standard measure of inequality in

 119  

health and in health care (29, 30). To measure the RCI we created a 2-dimensional graph

called a Concentration Curve (CC) by plotting cumulative percentages of the respective

outcome variables—eye screening service use or visual impairment—on the y-axis and

cumulative distributions of individuals by household income on the x-axis. The CC

allowed us to measure the distribution of the use of eye screening services or visual

impairment by aggregating across individuals for income rankings. The line of equity

(i.e., a 45-degree line) on the CC represents an equal distribution of the outcome variables

(i.e., 20%) across different income rankings. Based on the separate CC, we obtained the

RCIs of each of the outcomes—visual impairment and eye screening services. The RCI is

defined as twice the area between the CC and the line of equity and typically ranges

between -1 and 1. The RCIs for visual impairment and the use of eye care services were

calculated with the equation (1) (30, 31) :

𝑅𝐶𝐼 = !!!

ℎ!𝑟!   − 1 −  !!

!

!!!, (1)

Where ℎ! is the outcome, µ is the mean of the outcome, and 𝑟!  = i/N is the fractional rank

of individuals I in the median household income distribution, with i = 1 being the lowest

and i = N being the highest. If there is no income-related disparity (i.e., the CC is equal to

the line of equity), the RCI is zero. The convention is that the RCI takes a negative value

when the CC lies above the line of equity, indicating a disproportionate concentration of

health or health care among the lowest quintiles (i.e., “pro-poor”), and a positive value

when the CC lies below the line of equity (i.e., “pro-rich”). The greater the value of the

RCI, the greater the degree of concentration in a negative or positive direction.

After obtaining the RCIs for visual impairment and the use of eye screening services, we

applied the decomposition method proposed by Wagstaff et al. (32) to assess the major

contributors to income-related RCIs. The basic idea of decomposition is measuring

 120  

whether socio-demographic factors related to income such as age, sex, marital status,

educational attainment and complementary insurance coverage, etc. (i.e., explanatory

variables), are contributing to the overall level of income-related inequity (32, 33).

Assuming that the outcome variable is linked to k determinants through a linear

regression, it is possible to decompose the RCI by those k determinants. Decomposition

of the RCI is represented by equation (2):

𝑅𝐶𝐼 = (𝛽!𝑥!/! 𝜇)𝐶! +!!!!=  𝐶!   + 𝐺𝐶!/𝜇, (2)

Where µμ is the mean of the proportion of visual impairment or the use of eye screening

services, βk is the regression coefficients for the explanatory variables, 𝑥! is the mean of

individual explanatory variables, 𝐶! is income-related RCI, and 𝐺𝐶! is the generalized

concentration index for the error term. The overall disparity in visual impairment or eye

screening services has two components: (i) an explained component and (ii) unexplained

component. The explained component denotes the impact of socio-demographic

determinants on visual impairment or eye screening services and the extent of unequal

distribution of each determinant across different income groups. The residual component

reflects income-related disparities that are not explained by systematic variation of the

determinants in the model.

In addition to decomposing the RCIs, we calculated the horizontal inequity index (HI) for

the eye screening service models. HI represents the principle of equal treatment of people

in equal need, regardless of socio-demographic factors; it provides important evidence to

distinguish between inequity and inequality (34). Inequity is the result of systematic and

potentially remediable variation between population groups defined socially,

economically, or geographically and this is distinguished from inequality, which simply

indicates biological difference or individual’s choice (35). HI can be obtained by

 121  

subtracting the RCI of need-standardization from the total RCI as proposed by Wagstaff

et al. (32). A positive (negative) value of HI indicates horizontal inequity favoring the

better-off (worse-off); a zero value of HI means no horizontal inequity (i.e., eye screening

and need are proportionally distributed by need across income distribution). In

decomposition analysis for eye screening services, need factors for eye screening services

were represented through age, sex, and relevant health variables such as duration of

diabetes, and self-rated health, which reflect individual’s healthcare needs (36, 37). RCIs,

HIs, and the decomposition of RCIs were estimated using STATA 12 for MAC and

sampling weights provided by Statistics Canada and bootstrap method were applied in the

analyses.

 

5.3. Results

Descriptive statistics

We analyzed the data for 2,323 type 2 diabetic patients after removing missing data for

income and socio-demographic variables. Descriptive statistics for the sample, and the

population they represent, are presented in the previous chapter. Table 5.1 shows the

weighted percentage of visual impairment, eye screening and preventive eye screening

services according to self-reported household income quintiles. Patients with type 2

diabetes from lower income quintiles groups, in general, had a higher percentage of

visual impairment. The patterns of relationship between income and preventive eye

screening services, however, were different: only those in higher income groups used

more eye screening services.

Relative Concentration Index (RCI) and Horizontal inequity index (HI)

For visual impairment, a small, but pro-poor disparity was observed among type 2

diabetic patients (RCI = -0.104) (Table 5.2). This result indicates that more visual

 122  

impairment was concentrated in diabetic patients with lower income. There appeared to

be consistent, although small, pro-rich disparities in eye screening services and

preventive eye screening services among type 2 diabetic patients (Table 5.2), suggesting

that type 2 diabetic patients with higher incomes were more likely to receive eye

screening services. The magnitude of the RCI for preventive eye screening services,

which only included type 2 diabetic patients without visual impairment, was slightly

larger than the one for eye screening services in all type 2 diabetic patients (0.042 versus

0.025). However, the 95% confidence intervals overlap (Table 5.2).

Decomposition of the RCI

The decomposition analysis for eye screening services and preventive eye screening

services indicate that horizontal inequity exists. The observed RCI in both eye screening

services and preventive eye screening by income does not necessarily suggest “unfairness”

because of the underlying unequal distribution of need factors in the population (34).

After adjusting need factors that predict health care use including sex, age, duration of

diabetes, and self-rated health, HI still remains positive, indicating “pro-rich inequity in

the use of eye screening services and preventive eye screening services (Table 5.2).

Table 5.3-5.5 report the total RCI decomposition for the models referring to visual

impairment, eye screening services and preventive eye screening services, respectively.

The first column in each table shows elasticity for each determinant. The elasticity shows

a percentage of visual impairment or eye screening services due to a percentage change in

each determinant. Decomposition of concentration indices also showed the expected

distribution of determinants in income groups. Positive concentration index indicates

concentration of individuals with specific characteristics in higher income group.

 123  

Finally, the last three columns report, respectively, absolute, percentage and aggregated

contributions to total income-related disparities. Positive contribution of each socio-

demographic factor indicates that the factor is associated with both income and visual

impairment or eye screening services. In other words, positive contribution of the

determinant implies that the observed disparities can be x % reduced if the determinant

were to be distributed equally across income quintile groups, or if the determinants was

not associated with visual impairment or eye screening services. On the other hand,

negative contribution of a factor indicates contributing to a reduction in income-related

disparities.

The decomposition results show that age contributes 42% to the observed inequality

while income only contributes about 18% to the inequality in visual impairment among

type 2 diabetic patients (Table 5.3). The second greatest contributor was marital status.

These results demonstrate that the observed disparity in visual impairment would be 42%

and 25% lower respectively if patients who are older or live alone were equally

distributed across the income range or if being old and single had no effect on visual

impairment. Sex and duration of diabetes contributed respectively 9.5% and 5.3 % to the

observed inequality in visual impairment in type 2 diabetic patients (Table 5.3).

According to the decomposition results for eye screening services in type 2 diabetic

patients, the greatest contributions to inequity in the use of eye screening services came

from income, which explains 88 % of inequity in the use of eye screening services (Table

5.4). Having private insurance, discussion of complications with health professionals,

marital status and region of residence also contribute to the pro-rich disparities in the use

of eye screening services. The negative contribution from visual impairment indicates

contribute to a reduction in the observed pro-rich disparities.

 124  

The decomposition results for preventive eye screening services (Table 5.5) indicates that

the observed disparities can be explained by income (73.7 %), private insurance (42.8%)

and patients experience in discussion of diabetic eye complications with health

professionals (20.6%). In addition to these three main contributors, marital status and

regions of residence contributed to pro-rich disparities in the use of preventive eye

screening services.

5.4. Discussion  

Using the methods of RCI and decomposition of the RCI we measured income-related

inequities in both visual impairment and the use of eye screening services among type 2

diabetic patients living in Canada. For visual impairment, our results show similar trends

to that of general health; i.e., individuals with lower income tend to have more visual

impairment than individuals with higher income. Meanwhile, our decomposition analyses

for eye screening services indicate “pro-rich” disparities in the use of eye screening

services and preventive eye care services, suggesting that more eye screening services are

concentrated in type 2 diabetic patients with higher income groups.

The concentration of visual impairment in type 2 diabetic patients with lower

socioeconomic status (i.e., RCI negative, pro-poor direction), is not a new or surprising

finding. It is commonly believed that income plays a crucial role in determining health

(38, 39). For example, Canadians with the highest incomes are two and a half times more

likely to report good health than those with the lowest income (40). Similarly, those with

upper middle and middle income follow suit with likelihoods of reporting excellent or

very good health below their more affluent counterparts. There are various interpretations

how inequities in income are translated into health disparities (39-41). One of the

 125  

interpretations is that health disparities are caused by the differential accumulation of

exposures and experiences (39, 41). Under this interpretation, income-related health

disparity is an accumulative result of negative exposures and lack of material conditions

at the individual level, accompanied by systematic underinvestment across a broad range

of human, physical, health, and social infrastructure (39, 41). For example, it is well

acknowledged that income disparity in health is significantly associated with various

dimensions of infrastructure – unemployment, health insurance, social welfare, work

disability, educational and medical expenditure (39, 41).

The decomposition analysis for visual impairment showed that age and marital status

were the largest contributors to the existing pro-poor disparities, suggesting that most of

the income-related disparities in visual impairment are actually due to the concentration

of patients who are older or live alone in the lower tail of the income distribution.

However, these factors making large contributions may be difficult to be modified.

Firstly, greater age is known as one of risk factors for eye disease such as DR, glaucoma,

and cataracts and higher prevalence and severity of these eye complications increase in

older age groups (42-44). In addition, lower income is more concentrated in the older age

groups because many elderly patients may not be involved in the paid labour force, as

shown by the negative concentration index in Table 5.3. Lastly, compared to the younger

age group, there should be higher rate of older patients living alone. Considering the

major contributions of unmodified factors to overall disparity, the observed disparity in

visual impairment may be considered more of an inequality, rather than inequity. Both

inequality and inequity indicate differences in health or health care, the latter term implies

differences in health that is considered to be unfair or the result of injustice (45).

 126  

For the use of eye screening services, the RCI indicated a “pro-rich” direction, suggesting

that individuals with higher income were more likely to use regular eye screening

services. In the RCI model that only included type 2 diabetic patients with non-visual

impairment, we also observed a “pro-rich” direction, indicating that type 2 diabetic

patients with higher income were more likely to use preventive eye screening services,

even in the absence of diagnosed eye disease. In addition to the observed RCIs, our

results indicate that horizontal inequity, referring to unequal service for equal need, exists

for patients with diabetes living in Canada. This suggests that inequities in eye screening

services remain even after adjusting for need factors among type 2 diabetic patients.

Nonetheless, it should be recognized that the magnitude of the RCIs and HIs were

relatively small.

Previous studies of income-related inequities in health care using the RCI and

decomposition analysis have often found inequities in health care, with the main

contributor to the observed inequities being income (34, 46, 47). For instance, a Canadian

study found that income-related inequity in the use of health care services exists despite

Canada’s universal health care system, suggesting that while the current universal health

care system may help to reduce income-related barriers to health care, it fails to eliminate

them completely (47). Findings from recent research consistently show that use of

specialist services in Canada is more frequent for individuals with higher income (47).

McGrail et al. (2008) interpreted the observed inequities in specialist services in Canada

to be a result of physician referral patterns because there were no income-based

inequalities in the use of General Physician services (47). Studies have shown that a

patient’s preference or expectation plays an important role in explaining variation in the

use of specialist services between those of high and low SES (48, 49). The poor may be

less able to express their need for care. Furthermore, those of higher SES may have

 127  

different attitudes about the benefits that can be realised by accessing care (49).

Regarding use of eye screening services, it is plausible that lower income individuals may

not be as aware of the importance of regular screening services.

In our decomposition analysis, income had the largest contribution to disparities in both

eye screening and preventive eye screening services. Following income, private health

insurance for eye care was the second largest contributor to the existing inequities in the

use of these services. The finding suggests that under the current publicly funded health

care system in Canada, having private health insurance could be a crucial determinant in

the use of eye screening service because eye screening services by an optometrist may not

be covered by provincial health care plan. For example, prior to 2007, eye screening

services by an optometrist for diabetic patients was not reimbursed by the Alberta

provincial health care plan (2). This suggests that an increase in provincial coverage for

eye screening services could potentially reduce income-related pro-rich inequities.

Discussion of diabetic complications with health professionals was also an important

contributor to eye screening service utilisation. We found that discussion of diabetic

complications with a physician was associated with increased the use of eye care services.

A qualitative study suggested that detailed information from a primary care provider is a

pivotal factor in adherence to recommended eye screening services in diabetes

management (50). Because patients acquire the relevant information from their primary

care provider, primary care providers have an important role in reducing income-related

inequity in eye screening services.

Limitations

 128  

While the RCI, and decomposition of the RCI, are a common and widely used method for

quantifying the magnitude of inequities in health and health care, we have identified

several limitations. Firstly, decomposition analysis is not able to provide causal pathways

between the individual determinant and health/health care disparities. Secondly, the

decomposition method is a deterministic approach and there could be other factors, not

included in our model, that have may contributed to the observed inequities in visual

impairment and eye screening services. Although we included age, sex, duration of

diabetes, and self-rated health as need factors for eye screening services, there are many

other important clinical indicators of need for eye screening services, such as glycemic

control and blood pressure. Unfortunately these clinical data are not available with the

survey data. In addition, the SLCDC-DM 2011 is a self-reported survey data and

therefore prone to measurement errors and biases such as recall and social desirability

bias. Given that self-reported rates of regular eye screening services are higher than

previous Canadian studies (2, 18), social desirability is quite likely. Another potential

limitation is that sampling strategy which may limit the generalizability of findings; for

example, residents of three territories and of First Nations reserves were not included in

this survey.

5.5. Conclusion

Our findings suggest that disparities in visual impairment and the use of eye screening

services were observed among type 2 diabetic patients living in Canada, although these

disparities are small in magnitude. In particular, individuals with type 2 diabetes who

were of lower income are more likely to have visual impairment; however the magnitude

of the RCI is relatively small. The main contributors to the observed disparity were age

 129  

and marital status, suggesting that the observed disparity could be interpreted as

inequality rather than inequity.

Pro-rich disparities in both eye screening and preventive eye screening were found,

suggesting that patients with higher income are more likely to receive regular eye

screening services. The major contributors to the observed disparities were income,

private insurance and patient’s experience in discussion their complication with health

professionals. This finding suggests that income-inequity in the use of eye screening

services among type 2 diabetic patients may still exist under the universal health care

system despite the magnitude of inequities were relatively small. To develop health and

health care policy for tackling disparities in visual impairment and the use of eye

screening services, the identified determinants in our study could be considered.

 

 130  

Tables and Figures

Table 5.1. Visual impairment and eye screening services according household income

Income  quintile*  

Total  estimated    number  (N)  

Visual  impairment    

(%)  

Eye  screening    (%)  

Total  estimated    number  (N)  

Preventive  eye  screening    

(%)    

Lowest  quintile   96,125   54.0   71.0   44,183   59.2  

2nd  quintile   249,631   45.1   70.9   137,105   66.1  

3rd  quintile   334,876   31.6   69.8   228,894   62.0  

4th  quintile   341,034   27.1   69.8   248,752   62.5  

Highest  quintile     302,887   23.3   78.0   232,373   75.9  

Mean/Sum   1,324,553   32.7   71.9   891,307   65.6  

*income quintile groups were categorized based on the household income decile groups defined by Statistics Canada

Table 5.2. Measurement of disparities for visual impairment and eye screening services

Measurement  of  Disparity   Visual  impairment   Eye  screening    services  

Preventive    eye  screening  services  

Mean    (SE)   0.33    (0.02)   0.72  (0.15)   0.66  (0.21)  

Relative  Concentration  Index     -­‐0.104   0.025   0.043  (95%  CI)   (-­‐0.163,  -­‐0.046)   (-­‐0.001,  0.051)   (0.004,  0.082)  

Horizontal  Inequity  Index  (HI)       0.039   0.056  

 131  

Table 5.3. Decomposition results for visual impairment among type 2 diabetic patients

    Elasticity   CI   Contribution   Total    (%)  

Aggregated  contribution  (%)  

Sex  -­‐  Female   0.083   -­‐0.120   -­‐0.010   9.5   9.5  

Age  50-­‐64   0.129   0.116   0.015   -­‐14.3   42.0  

Age  65-­‐79   0.348   -­‐0.094   -­‐0.033   31.5    Age  80  or  more   0.121   -­‐0.215   -­‐0.026   24.8    Duration  of  DM   0.188   -­‐0.029   -­‐0.005   5.3   5.3  

Marital  status   0.142   -­‐0.188   -­‐0.027   25.6   25.6  

Lower  income   0.051   -­‐0.371   -­‐0.019   18.2   18.2  

Lower  education   -­‐0.072   -­‐0.182   0.013   -­‐12.6   -­‐12.6  

Official  language  at  home   -­‐0.126   -­‐0.425   0.011   -­‐10.2   -­‐10.2  

Immigrant   -­‐0.009   -­‐0.203   0.002   -­‐1.7   -­‐1.9  

Aboriginal   0.004   0.056   0.000   -­‐0.2    Quebec   0.025   -­‐0.148   -­‐0.004   3.6   -­‐3.9  

Ontario   0.111   0.023   0.003   -­‐2.5    Prairie   0.019   0.074   0.001   -­‐1.3    BC   0.037   0.104   0.004   -­‐3.7    

Urban/Rural   0.010   -­‐0.050   -­‐0.001   0.5   0.5  

Private  insurance   0.043   0.057   0.002   -­‐2.4   -­‐2.4  

Discussion  with  HP   0.151   0.033   0.005   -­‐4.8   -­‐4.8  

Sum               65.4      

Residual               34.6      

 132  

Table 5.4. Decomposition results for eye screening services among type 2 diabetic patients

Decomposition   Elasticity   CI   Contribution   Total  (  %)  

Aggregated  contribution  (%)  

Sex-­‐  Female   0.021   -­‐0.120   -­‐0.003   -­‐10.2   -­‐10.2  

Age  50-­‐64   -­‐0.018   0.116   -­‐0.002   -­‐8.3   -­‐3.7  

Age  65-­‐79   -­‐0.004   -­‐0.094   0.000   1.4    Age  80  or  more   -­‐0.004   -­‐0.215   0.001   3.2    Duration  of  DM   0.048   -­‐0.029   -­‐0.001   -­‐5.7   -­‐5.7  

Visual  impairment   0.083   -­‐0.088   -­‐0.007   -­‐29.8   -­‐29.8  

Self-­‐health-­‐  Good   -­‐0.028   0.031   -­‐0.001   -­‐3.6   -­‐8.3  

Self-­‐health-­‐  Fair   -­‐0.020   -­‐0.102   0.002   8.5    Self-­‐health-­‐  Poor   0.010   -­‐0.313   -­‐0.003   -­‐13.2    

Sub-­‐Sum           -­‐0.014   -­‐57.7   -­‐57.7  

Need  factor  CI       -­‐0.014              

           Marital  status   -­‐0.018   -­‐0.188   0.003   13.6   13.6  

Lower  income   -­‐0.059   -­‐0.371   0.022   88.2   88.2  

Lower  education   0.000   -­‐0.182   0.000   -­‐0.1   -­‐0.1  

Official  language  at  home   0.003   -­‐0.425   -­‐0.001   -­‐5.2   -­‐5.2  

Immigrant   0.003   -­‐0.203   -­‐0.001   -­‐2.4   -­‐3.0  

Aboriginal   -­‐0.003   0.056   0.000   -­‐0.6    Quebec   -­‐0.009   -­‐0.148   0.001   5.1   7.7  

Ontario   -­‐0.002   0.023   0.000   -­‐0.2    Prairie   0.009   0.074   0.001   2.6    BC   0.001   0.104   0.000   0.2    

Urban/Rural   0.020   -­‐0.050   -­‐0.001   -­‐4.0   -­‐4.0  

Private  insurance   0.213   0.057   0.012   49.3   49.3  

Discussion  with  HP   0.151   0.033   0.005   20.1   20.1  

Sub-­‐sum           0.041   166.6      

Non-­‐need  factor  CI       0.041              

Residual           -­‐0.002   -­‐8.9      

HI  (RCI-­‐Need  CI)       0.039              

 

 133  

Table 5.5. Decomposition results for preventive eye screening services among type 2 diabetic

patients

    Elasticity   CI   Contribution   Total    (%)  

Aggregated  contribution  (%)  

Sex-­‐  Female   0.026   -­‐0.111   -­‐0.003   -­‐6.7   -­‐6.7  

Age  50-­‐64   -­‐0.033   0.100   -­‐0.003   -­‐7.8   -­‐12.1  

Age  65-­‐79   0.015   -­‐0.099   -­‐0.001   -­‐3.4    Age  80  or  more   0.001   -­‐0.295   0.000   -­‐0.8    Duration  of  DM   0.042   -­‐0.013   -­‐0.001   -­‐1.2   -­‐1.2  

           Self-­‐health-­‐  Good   -­‐0.033   -­‐0.004   0.000   0.3   -­‐11.1  

Self-­‐health-­‐  Fair   -­‐0.022   -­‐0.057   0.001   3.0    Self-­‐health-­‐  Poor   0.016   -­‐0.384   -­‐0.006   -­‐14.4    

Sub-­‐Sum           -­‐0.013   -­‐31.2   -­‐31.2  

Need  factor  CI       -­‐0.013              

           Marital  status   -­‐0.026   -­‐0.144   0.004   8.6   8.6  

Lower  income   -­‐0.083   -­‐0.382   0.032   73.7   73.7  

Lower  education   0.015   -­‐0.162   -­‐0.002   -­‐5.8   -­‐5.8  

Official  language  at  home   0.000   -­‐0.446   0.000   -­‐0.5   -­‐0.5  

Immigrant   0.007   -­‐0.253   -­‐0.002   -­‐4.1   -­‐4.5  

Aboriginal   -­‐0.002   0.069   0.000   -­‐0.3    Quebec   -­‐0.017   -­‐0.130   0.002   5.3   11.6  

Ontario   0.000   -­‐0.019   0.000   0.0    Prairie   0.014   0.155   0.002   5.0    BC   0.004   0.145   0.001   1.3    

Urban/Rural   0.025   -­‐0.041   -­‐0.001   -­‐2.4   -­‐2.4  

Private  insurance   0.254   0.072   0.018   42.8   42.8  

Discussion  with  HP   0.042   0.210   0.009   20.6   20.6  

Sub-­‐sum           0.062   144.2      

Non-­‐need  factor  CI       0.062              

Residual           -­‐0.006   -­‐13.1    HI  (RCI-­‐Need  CI)       0.056              

 134  

Figure 5.1. Aggregated contributions to RCI for visual impairment

 

 

 

 

-­‐20.0   -­‐10.0   0.0   10.0   20.0   30.0   40.0   50.0  

Sex  Age  

Duration  of  DM  Marital  status  

Income  Education  

OfDical  language  at  home  Demographic  

Provinces  Residence  

Private  insurance  Discussion  with  HP  

Contribution  to  the  RCI  for  visual  impairment  

 135  

Figure 5.2. Aggregated contributions to RCI for eye screening services

 

 

 

 

 

-­‐40.0   -­‐20.0   0.0   20.0   40.0   60.0   80.0   100.0  

Sex  Age  

Duration  of  DR  Visual  Impairment  

Self  health  Marital  status  

Income  Education  

Demographic  Provinces  

OfDical  language  at  home  Residence  

Private  insurance  Discussion  with  FP  

Contribution  to  the  RCI  for  eye  screening    

 136  

Figure 5.3. Aggregated contributions to RCI for preventive eye screening services

-­‐20   -­‐10   0   10   20   30   40   50   60   70   80  

Sex  Age  

Duration  of  DM  Self  health  

Marital  status  Income  

Education  Demographic  

Provinces  OfDical  language  at  home  

Residence  Private  insurance  Discussion  with  HP  

Decomposition  of  the  RCI  for  preventive  eye  screening  services  

 137  

References  

1. George M, Harper R, Balamurugan A, Kilmer G, Bynum L. Diabetic Retinopathy and Its Risk Factors in a Population-Based Study. J Prim Care Community Health. 2011;2(2):122-6. 2. Rudnisky CJ, Tennat MTS, Johnson JA, Balko SU. Chapter 9 : Diabetes and Eye Disease in Alberta. In: Johnson JA,editor Alberta Diabetes Atlas 2011. Edmonton: Institute of Health Economics; 2011. p. 153-74. 3. Public Health Agency of Canada. Complications of Diabetes. Ottawa: Public Health Agency of Canada,2008; [http://www.phac-aspc.gc.ca/cd-mc/diabetes- diabete/diabetes_complications-diabete_complications-eng.php]. 4. Perruccio AV, Badley EM, Trope GE. A Canadian population-based study of vision problems: assessing the significance of socioeconomic status. Can J Ophthalmol. 2010;45(5):477-83. 5. American Academy of Ophthalmology. Chapter 5 : Retinal Vascular Disease. Retina and Vitreous Section12 Basic and Clinical Science Course 2011-2012. San Fransico, CA: American Academy of Ophthalmology The Eye M.D. Association; 2011. p. 107-80. 6. Hooper P, Boucher MC, Cruess A, Dawson KG, Delpero W, Greve M, et al. Canadian Ophthalmological Society evidence-based clinical practice guidelines for the management of diabetic retinopathy. Can J Ophthalmol. 2012;47(2 Suppl):S1-30. 7. Bond SR, Advani A, Altomare F., Stockl F. Canadian Diabetes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada: Retinopathy. Can J Diabetes. 2013;37(suppl1):S137-S41. 8. van Eijk KN, Blom JW, Gussekloo J, Polak BC, Groeneveld Y. Diabetic retinopathy screening in patients with diabetes mellitus in primary care: Incentives and barriers to screening attendance. Diabetes Res Clin Pract. 2011. 9. Buhrmann R, Houdege W, Beardmore J, Baker G, Lowcock B, Pan I, Bovell A. Chapter 2. The Epidemilogoy of Blindness and Visual Imparment in Canada : Diabetic Retinopathy. Foundations for a Canadian Vision Health Strategy : Towards Preventing Avoidable Blindness and Promoting Vision Health: The National Coalition for Vision Health; 2007. p. 14-5. 10. Public Health Task Force for Vision and Ophthalmology. Determining the information needs for prevention of vision loss strategies in Canada. Public Health Task Force for Vision and Ophthalmology. Can J Ophthalmol. 2000;35(5):255-7. 11. Szabo SM, Beusterien KM, Pleil AM, Wirostko B, Potter MJ, Tildesley H, et al. Patient Preferences for Diabetic Retinopathy Health States. Invest Ophthalmol Vis Sci. 2010;51(7):3387-94. 12. Aiello LP, Cavallerano J, Bursell SE. Diabetic eye disease. Endocrinol Metab Clin North Am. 1996;25(2):272-90.

 138  

13. Kline R, Kline B.. Chapter 14. Visual disorders in diabetes. Diabetes in American, 2nd edition. Bethesda:National Institute of Health;1995. p. 293 -338. 14. McCarty CA, Lloyd-Smith CW, Lee SE, Livingston PM, Stanislavsky YL, Taylor HR. Use of eye care services by people with diabetes: the Melbourne Visual Impairment Project. Br J Ophthalmol. 1998;82(4):410-4. 15. Gulliford MC, Dodhia H, Chamley M, McCormick K, Mohamed M, Naithani S, et al. Socio-economic and ethnic inequalities in diabetes retinal screening. Diabet Med. 2010;27(3):282-8. 16. Millett C, Dodhia H. Diabetes retinopathy screening: audit of equity in participation and selected outcomes in South East London. J Med Screen. 2006;13(3):152-5. 17. Sanmartin C, Gillmore J. Diabetes care in Canada: Results from selected provinces and territory, 2005. Ottawa: Statistics Canada, 2006. 18. Burhrmann R, Assaad D, Hux JE, Tang M, & Sykora K. Chapter 10: Diabetes and the Eye. Diabetes in Ontario: An ICES Practice Atlas: Institute for Clinical Evaluative Sciences; 2003. p. 10.193- 10.208. 19. Brown AF, Ettner SL, Piette J, Weinberger M, Gregg E, Shapiro MF, et al. Socioeconomic Position and Health among Persons with Diabetes Mellitus: A Conceptual Framework and Review of the Literature. Epidemiol Rev. 2004;26(1):63- 77. 20. Moss SE, Klein R, Klein BEK. Factors Associated With Having Eye Examinations in Persons With Diabetes. Arch Fam Med. 1995;4(6):529-34. 21. Saadine JB, Fong DS, Yan J. Factors Associated With Follow-Up Eye Examinations Among Persons With Diabetes. Retina. 2008;28(2):195-200 1. 22. Fraser S, Edwards L. Health Equity Audit - Diabetic Retinopathy Screening : Southampton City, Hampshire, Isle of Wight and Portsmouth City PCTs. Southampton City PCT, July 2010. 23. Statistics Canada. Survey on Living with Chronic Diseases in Canada User guide. Ottawa; Canada: Statistics Canada, 2011. 24. Ng E VS, Geiss L, Johnson JA. Concordance between self-report and a survey-based algorithm for classification of type 1 and type 2 diabetes using the 2011 population- based Survey on Living with Chronic Diseases in Canada (SLCDC) Diabetes Component. Can J Diabetes. 2013;Forthcoming. 25. Statistics Canada. Canadian Community Health Survey (CCHS) Annual Component - Public Use Microdata File, 2009-2010 Dervied Variable (DV) Specifications. Ottawa: Statistics Canada, 2011. P.82-6. 26. Dinca-Panaitescu S, Dinca-Panaitescu M, Bryant T, Daiski I, Pilkington B, Raphael D. Diabetes prevalence and income: Results of the Canadian Community Health

 139  

Survey. Health Policy. 2011;99(2):116-23. 27. Kakwani N. Income Inequality and Poverty : Methods of Estimation and Policy Applications. New York: Oxford University Press; 1980. 28. Wagstaff A, Paci P, van Doorslaer E. On the measurement of inequalities in health. Soc Sci Med. 1991;33(5):545-57. 29. O'Donnell O, van Dooslaer E, Wagstaff A, Lindelow M. Chapter 7. Concentration Curves. Analyzing Health Equity Using Household Survey Data : A Guide to Techniques and Their Implementation. Washington, DC: The World Bank; 2009. p. 83-94. 30. O'Donnell O, van Dooslaer E, Wagstaff A, Lindelow M. Chapter 8. The Concentration Index. Analyzing Health Equity Using Household Survey Data : A Guide to Techniques and Their Impementation. Washington, DC: The World Bank; 2008. p. 95-108. 31. Haper S, Lynch J. Methods for Measuring Cancer Disparities : Using Data Relevant to Healthy people 2010 Cancer-Related Objectives. Bethesda, MD: National Cancer Institute; 2005. 32. Wagstaff A, van Doorslaer E, Watanabe N. On decomposing the causes of health sector inequalities with an application to malnutrition inequalities in Vietnam. J Econom. 2003;112(1):207-23. 33. O'Donnell O, van Dooslaer E, Wagstaff A, Lindelow M. Chapter 13. Explanning Socioeconomic-Related Health Inequality : Decomposition of the Concentration Index. Analyzing Health Equity Using Household Survey Data : A Guide to Techniques and Their Impementation. Washington, DC: The World Bank; 2008. p. 159-64. 34. Sara A. Does Equity in Healthcare Use Vary across Canadian Provinces? Healthcare Policy. 2008;3(4):83-99. 35. Starfield B. The hidden inequity in health care. Int J Equity Health. 2011;10(1):15. 36. Watanabe R, Hashimoto H. Horizontal inequity in healthcare access under the universal coverage in Japan; 1986–2007. Soc Sci Med. 2012;75(8):1372-8. 37. van Doorslaer E, Wagstaff A, Bleichrodt H, Calonge S, Gerdtham U-G, Gerfin M, et al. Income-related inequalities in health: some international comparisons. J Health Econ. 1997;16(1):93-112. 38. Marmot M. The Marmot review final report: fair society, healthy lives. 2010. University College London. 39. Lynch J, Smith GD, Harper S, HillemeireM, Ross N, Kaplan GA, Wolfson M. Is Income Inequality a Determinant of Population Health? Milban Q. 2004;82(1):5-99. 40. Frohlich KL, Ross N, Richmond C. Health disparities in Canada today: some

 140  

evidence and a theoretical framework. Health Policy. 2006;79(2-3):132-43. 41. Lynch JW, Smith GD, Kaplan GA, House JS. Income inequality and mortality: importance to health of individual income, psychosocial environment, or material conditions. BMJ. 2000;320(7243):1200-4. 42. Williams R, Airey M, Baxter H, Forrester J, Kennedy-Martin T, Girach A. Epidemiology of diabetic retinopathy and macular oedema: a systematic review. Eye. 2004;18(10):963-83. 43. Klein BEK. Overview of Epidemiologic Studies of Diabetic Retinopathy. Ophthalmic Epi. 2007;14(4):179-83. 44. O'Donnell K. Eye care in the UK : Epidemiology, intervention and ethnicity. Public Health Action Support Team, August 2009. 45. Kawachi I, Subramanian SV, Almeida-Filho N. A glossary for health inequalities. J Epidemiol Commun H. 2002;56(9):647-52. 46. van Doorslaer E. Masseria C. Income-Related Inequality in the Use of Medical Care in 21 OECD Countries. OECD health working paper. Paris: Organization for Economic Co-operation and Development. May, 2004. 47. McGrail KM. Income-related inequities: Cross-sectional analyses of the use of medicare services in British Columbia in 1992 and 2002. Open Med. 2008;2(4): e3- 10. 48. Langley GR, Maclellan AM, Sutherland HJ, Till JE., Effect of non-medical factors on family physicians' decisions about referral for consultation. CMAJ. 1992;147:659-66. 49. Dunlop S, Coyte PC, McIsaac W. Socio-economic status and the utilisation of physicians' services: results from the Canadian National Population Health Survey. Soc Sci Med. 2000;51(1):123-33. 50. Yeo ST, Edwards RT, Luzio SD, Charles JM, Thomas RL, Peters JM, et al. Diabetic retinopathy screening: perspectives of people with diabetes, screening intervals and costs of attending screening. Diabet Med. 2012;29(7):878-85.

 141  

Chapter 6. Conclusion

6.1. Summary of Research

The findings from this program of research suggest that socioeconomic-related disparities

in visual impairment and eye screening services in patients with diabetes exist in Canada.

The results of three separate analyses indicated socioeconomic disparities in eye care

services at the provincial level and income-related disparities in visual impairment and

eye screening services at the national level. However, while they exist, and are

measureable with the Relative Concentration Index (RCI) method, the magnitudes of

these disparities are quite small when considered at the population level.

Nonetheless, when we explored variations in the disparities using a decomposition

analysis of the RCI, a number of important trends were apparent. At the provincial level,

the observed patterns of disparity in eye care services varied by socioeconomic indicator:

household income and material deprivation indices consistently showed a “pro-rich”

pattern, while the social deprivation index indicated a “pro-poor” pattern. Our findings

also suggested that material deprivation index and place of residence (urban/rural) were

important contributors to the observed “pro-rich” income- and material deprivation index-

related disparities. The social deprivation-related disparity was explained largely by

social-deprivation itself. At the national level, like at the provincial level, income-related

disparities in eye screening services and preventive eye screening services revealed a

“pro-rich” pattern while the disparity in visual impairment indicated a “pro-poor” pattern.

The main contributor to the observed disparities in both eye screening service and

preventive eye screening was income while the disparity in visual impairment was

predominantly related to age.

 142  

In addition, an examination of socioeconomic factors associated with visual impairment

and both general eye screening and preventive eye screening services among Canadians

living with diabetes provided further evidence that demographic factors such as age, sex,

and duration of diabetes were associated with visual impairment. Regarding eye

screening services and preventive eye screening services, in addition to household income,

factors such as patient’s experience in discussing diabetic eye complications with health

professionals and having private insurance covering eye care appointment were

associated with more regular eye screening services among diabetic patients.

The observed disparities in eye examination in diabetic populations from this line of

research are consistent with findings from health disparity research in the U.K., where a

similar publicly funded health care system exists (1-4). For instance, Gulliford and his

colleagues reported that diabetic patients residing in most deprived areas were more than

1.4 times as less likely to have received eye screening services compared to those

residing in least deprived areas (1). Also, Hippisley-Cox et al. stated that patients in areas

of high deprivation were less likely to have retinal screening compared with those from

affluent areas (4). These previous findings of disparities in eye examinations were mainly

assessed by logistic regression-based analyses, and represented by odds-ratios (ORs).

These results can be interpreted as a relative comparison, showing the proportional

increase or decrease of relative risk in use of eye examination for a one-unit change in

socioeconomic group (5). Meanwhile, our results of disparities from the RCI analyses

indicate positive or negative value of the RCI, showing a disproportionate share of eye

examination across socioeconomically ranked groups (5).

6.2. Strengths, Significance and Implication of Research

 143  

Diabetic retinopathy (DR) has emerged as an important health care priority as DR is the

leading cause of blindness and visual impairment among patients with diabetes (6). It has

also been estimated that nearly half a million Canadians currently have some form of DR,

and approximately 100,000 Canadians have PDR (a more advanced form of DR), diabetic

macular edema or both (7). Glaucoma, cataracts, and other disorders of the eye also occur

earlier and more frequently in people with diabetes (8). Vision impairment caused by DR

can be prevented and more easily managed when the disease is detected and treated early

in its course (9-11). Accumulated research suggests that regular eye examination can help

to decrease the risk of severe vision loss by >90%, by providing early detection and

timely treatment (6, 12). Periodic eye screening also substantially contributes to cost

savings. Previous studies have shown that the cost of eye screening services is much less

than the costs associated with providing social support for visual impairment (13-15).

Partly for this reason, clinical practice guidelines for DR screening established by the

Canadian Diabetes Association (CDA) recommend that all diabetic patients receive an

annual eye examination (10).

Despite the importance of regular eye examinations for preventing visual impairment in

diabetic patients, there is limited Canadian evidence on factors associated with visual

impairment and eye screening services in the diabetic population. Consequently, both

federal and provincial governments and policy makers may remain unaware of the current

magnitude and, more importantly, the potential for rises in diabetic eye diseases and the

resulting need for regular eye screening services (8).

To our knowledge, this program of the research, using provincial population datasets and

a nationally representative survey dataset, is the first attempt to measure and understand

 144  

socioeconomic disparities focusing on visual impairment and eye screening services in

Canadian patients with diabetes. It is also the first study to identify socioeconomic factors

associated with visual impairment and eye screening services in Canada. Our program of

research provides evidence on previously unexplored issues in disparities in health and

health care in Canadians living with diabetes. While there is a bourgeoning literature on

health and health care disparities, most previous studies are based on a macro study of

inequity in health or health care. The line of our research focused particularly on visual

impairment and eye screening services among diabetic patients, and our micro-level

investigation of a specific disease and service category benefits to address the issue of

appropriateness or quality in particular health care services (16).

Using provincial surveillance datasets for the past 15 years, we were able to obtain

relatively precise estimates and to observe the patterns of the disparities. One of the

limitations of current literature is the use of cross-sectional data, which make it difficult

to see the pattern of health and health care disparities (17). Our findings from the Alberta

Diabetes Surveillance System (ADSS), using data from 1995-2009, allowed us to observe

changes in disparities across different socioeconomic indicators and infer how changes of

provincial health policy might impact on changes in health and health care disparities in

diabetic patients living in Alberta.

In addition, to enhance our ability to understand disparities in eye care services at the

provincial level, we linked the ADSS, the provincial surveillance dataset, to different

socioeconomic indicators. The ADSS datasets includes household income and the

Canadian Deprivation Index, which are derived from the 2006 Canada Census.

Household income, reflecting material goods and services, is widely used in research and

 145  

considered a standard measure of socioeconomic status (18). Nevertheless, while income

is a “gold standard” indicator of health and health care and has ramifications for other

determinants of health, it cannot take the place of the other determinants (19). To take the

multidimensional concept of socioeconomic status into account, deprivation indices were

developed and are widely used in the U.K., Europe, and Canada (20). The material

deprivation index used in our study represents the level of educational attainment (i.e.,

percentage of secondary school completion), employment status, and income (20). The

social deprivation index mainly reflects the prevalence of single-parents families, of

people living alone, and of those who are separated, divorced or widowed (20).

Our findings suggest that economic indicators, represented by income and material

deprivation, indicate “pro-rich” disparities in use of eye care services, with a slightly

larger magnitude of material deprivation-related disparities compared to income-related

disparities. This finding implies that material deprivation index might capture more

economic or different dimension of economic-related factors than does income alone.

Previous study has documented that material deprivation, measured by the economic

exclusion index, is not captured by income alone, suggesting that income may be an

insufficient indicator of socioeconomic condition (21).

On the other hand, when based on the social deprivation index, disparities were in the

“pro-poor” direction: the opposite indicated by income- or material deprivation-related

RCI. This counterintuitive finding might be explained by previous finding (22),

suggesting that retired patients might use more health care services because they would

focus more time and attention on their health. Alternatively, the fact that there are more

socially deprived areas in urban setting, where more ophthalmology services are available,

could help explain the “pro-poor” direction of social deprivation-related RCI. In our

 146  

descriptive analysis of the distribution of social deprivation ranks across the province, the

lower social deprivation ranks were more concentrated in Edmonton and Calgary. It is

therefore plausible that a large number of individuals residing in more socially deprived

neighbourhood in urban areas could more easily access eye care services, and this may

result in a “pro-poor” direction of social deprivation-related RCIs. As a result, social

deprivation may not be a strong predictor of the use of eye care services. In fact, it has

been suggested that all component of social capital may not be related to use of health

care services due to the complex nature of social capital (23).

The RCI and decomposition analysis are now widely accepted by international

organizations such as the Organisation for Economic Co-operation and Development and

World Bank in studies measuring socioeconomic disparities (24-28). In addition, these

analyses are considered a standard quantitative measure of health and health care

disparities and can be used in studies to provide policy evidence to alleviate existing

health and health care disparities (5, 29). We adopted these methods for our research to

assess disparities in eye care services and visual impairment. The concentration index has

proven to be a useful summary statistic in the measurement of socioeconomic disparities

in the health sector and is accompanied by a relatively straightforward graphical

representation of disparities. This simplicity allows the comparison of different time

periods and facilitates an understanding of disparities by policy makers and other

stakeholders. Although using the RCI method strengthens our research, our interpretation

of the results focused on the magnitude of the RCI. In general, the observed degree of the

RCI can be interpreted based on absolute values (16, 24, 30-33); however, the value of

the RCI from the line of our research is relatively small. A number of Canadian studies

have reported similarly small magnitude of disparities in specialist care (16, 32-34). In

addition, the studies from the U.K. and several European countries have also exhibited

 147  

disparities in health and health care of similar magnitudes (24, 35). These common

findings support our conclusion that the magnitude of socioeconomic-disparities in health

and health care within a publicly funded health care system could be very small,

suggesting that socioeconomic barriers to health care within such a system are alleviated,

but are not completely eliminated.

To set a health policy priority, it may be meaningful to compare the RCIs for different

health issues. For example, a recent Canadian study on oral health that also used the RCI

method based on the same CCHS data and self-reported household income, reported a

magnitude of the RCI (0.26) larger than the RCI for we observed for visual impairment in

diabetic patients (0.11) (36). This direct comparison might be taken to mean that the

disparity in oral health in the general population may be a bigger priority than the

disparity in eye care services or visual impairment for diabetic patients in Canada.

Nonetheless, as with most policy discussions, there are other considerations that should

be taken into account in priority setting, including the absolute number of individuals

affected, and the economic considerations in establishing strategies to ameliorate the

disparities.

This line of research program was initiated based on integrated knowledge translation

approach in order to support an identified knowledge gap of concern to the clinical

community. The underuse of recommended eye examination was initially identified as an

increasing concern in Alberta Diabetes Atlas 2011 (9). Dr. Chris Rudnisky, a co-author of

the Alberta Diabetes Atlas 2011 was engaged to investigate this line of research program

in addition to Dr. Jeffrey Johnson and Dr. Sara Bowen. The collaboration with Dr.

Rudnisky has enriched our research capacity, particularly in understanding provision of

 148  

eye care services in real clinical settings in addition to clinical information on eye

complications among patients living with diabetes.

6.3. Limitations, and Implications for Future Research

Although we were able to add to the growing evidence assessing socioeconomic-related

disparities in eye screening services and visual impairment using a state-of-the-art

econometric approach, there were a number of limitations to our research that further

studies could address.

Firstly, using both provincial administrative data and national survey data limits our

ability to accurately capture the existing socioeconomic disparities due to some degree.

For instance, provincial administrative data allows us to identify all individuals diagnosed

with diabetes and their use of ophthalmology services, but individuals who have not

accessed the provincial health care services are not included, potentially underestimating

the magnitude of socioeconomic disparities in the use of eye care services among

individuals with lower socioeconomic position.

In addition, the provincial datasets did not include eye-screening services by optometrists,

because provincial health policy did not allow reimbursement of costs for these services

until the end of 2007. Yet, as reflected in the current Canadian Diabetes Association

guidelines, diabetic patients should receive eye screening services for DR from an

experienced eye care specialist, including both ophthalmologists and optometrists (9, 37).

The lack of data on eye screening services by optometrists may bias our estimate of

disparities in the use of eye screening services, exaggerating the degree of the disparities.

For example, the exclusion of patients who used optometry services might have led to

overestimation of the RCI and horizontal inequity index in chapter 2 and 3. Since eye

 149  

screening services by optometrists has been covered by the Alberta’s provincial health

care plan since 2007 (9), future research should aim to capture the services provided by

optometrists in order to accurately measure disparities in eye screening services in the

diabetic population. Regarding the SLCDC data, a nationally representative survey data,

we observed that the rate of eye screening was higher than the rate from our provincial

data. The higher rate of eye screening services in the SLCDC data may simply reflect the

fact that individual were asked if they received an eye screening service by either

ophthalmologists and optometrists while the provincial data only capture the eye

screening services by ophthalmologists. However, it needs to be considered that the

SLCDC is a survey data that may have patient’s recall bias and non-response bias.

Moreover, as we included all services provided by ophthalmologists in the analysis of

data from the ADSS, we did not differentiate screening from treatment, hence we referred

to disparities in “eye care services”. In our analysis of the SLCDC data, we attempted to

differentiate preventive eye screening services by limiting our analyses to those

individuals reporting no visual impairment. Future research using administrative data and

looking separately at optometrist and ophthalmologists’ services may be better suited to

differentiate screening and treatment, respectively.

Secondly, we were unable to determine if the observed disparities in eye screening

services were a result of access or utilization using the national survey data. While

“access” and “utilization” are used interchangeably, these terms should be distinguished

(38, 39). Access is a matter of supply only, utilization is a function of both supply and

demand (39). As such, equity of access can be achieved by providing some services to

patients rather than taking the patients to the services (39). In contrast, utilization of

healthcare services implies not only access but also an individual’s perception of the

 150  

benefits of health care (39). For instance, individuals may not fully understand the

benefits of health care, and fail to use the services although they can easily access the

services. This subject requires further attention and quantitative and qualitative methods

could be used to answer this question.

Thirdly, being limited to using provincial health care administrative datasets and a

national survey data, we were unable to determine individual’s clinical need and

appropriateness of the eye care and eye screening services received or not received. In

our decomposition analyses and logistic regression, we only included age, sex, duration

of diabetes and self-rated health (the SLCDC analyses only) as predictors of eye health

and eye care (screening) services in diabetic patients. However, there could be other

important clinical indicators associated with eye health and eye care (screening) services.

Unfortunately, our data were limited to the administrative database and the self-reported

survey dataset, which lack information on these important clinical parameters. For a

better understanding of disparities in diabetic population, future study needs to use data

including clinical information.

Finally, we were not able to fully explain the opposing patterns of the RCI for material

deprivation and social deprivation in the analyses based on ADSS data. One possible

explanation for the different direction of social deprivation-related RCI is the

independence of the material and social deprivation indices. The material and social

deprivation indices were constructed with principal component analysis (PCA) and

applying a varimax rotation in order to improve readability and to make these two factors

independent (20). The scoring used for PCA results in the material deprivation index and

social deprivation index being uncorrelated (20). Still, the lack of correlation does not

explain the opposite relationship between the material deprivation- and social

 151  

deprivation-related RCI over time in the province as a whole. Further study is required to

understand the opposite relationship between these two different deprivation-related

disparities.

6.4. Conclusions

Identifying the roots of inequities is a primary goal of public health. It is imperative that

action on the causal factors causing inequities be undertaken. Our research results provide

Canadian evidence on disparities in visual impairment and eye screening services in

diabetic patients at both provincial and national levels. Through two separate analyses

using provincial administrative datasets, we observed “pro-rich” economic disparities in

use of eye care services in Alberta, and “pro-poor” social disparities in eye care services

in Alberta. These disparities were explained mainly by the material deprivation and social

deprivation index, respectively. In our analyses using a national survey data, we also

observed a “pro-rich” direction of income-related disparities in both eye screening

services and preventive eye screening services. In addition, we found a “pro-poor”

direction of income-related disparity in visual impairment among type 2 diabetic patients.

The main contributors to the observed income-related disparity in visual impairment were

demographic factors such as age, sex, and marital status. For income-related disparities in

eye screening services and preventive eye screening services, enabling factors in addition

to income were the patient’s experience in discussing diabetic complication with a health

professional and having private health insurance coverage for eye care appointments. To

understand causal pathways on this topic, further research is required using both

provincial and national datasets as well as using qualitative methods.

 152  

References

1. Gulliford MC, Dodhia H, Chamley M, McCormick K, Mohamed M, Naithani S, et al. Socio-economic and ethnic inequalities in diabetes retinal screening. Diabet Med. 2010;27(3):282-8. 2. Chaturvedi N, Stephenson J, Fuller J. The Relationship Between Socioeconomic Status and Diabetes Control and Complications in the EURODIAB IDDM Complications Study. Diabetes care. 1996;19(5):423-30. 3. Moss SE, Klein R, Klein BEK.. Factors Associated With Having Eye Examinations in Persons With Diabetes. Arch Fam Med. 1995;4(6):529-34. 4. Hippisley-Cox J, O'Hanlon S, Coupland C. Association of deprivation, ethnicity, and sex with quality indicators for diabetes: population based survey of 53,000 patients in primary care. BMJ. 2004;329(7477):1267-9. 5. Haper S, Lynch J. Methods for Measuring Cancer Disparities : Using Data Relevant to Healthy people 2010 Cancer-Related Objectives. Bethesda: National Cancer Institute; 2005. 6. Garg S, Davis R. Diabetic Retinopathy Screening Update. Clin Diabetes. 2009;27(4):140-5. 7. Buhrmann R, Hodeg W, Beardmore J, BakerG, LowcockB, Pan I, BovellA. Section 2. The Epidemilogoy of Blindness and Visual Imparment in Canada : Diabetic Retinopathy. Foundations for a Canadian Vision Health Strategy : Towards Preventing Avoidable Blindness and Promoting Vision Health: The National Coalition for Vision Health; 2007. p. 12-26. 8. Amos AF, McCarty DJ, Zimmet P. The rising global burden of diabetes and its complications: estimates and projections to the year 2010. Diabet Med. 1997;14 Suppl 5:S1-85. 9. Rudnisky CJ, Tennat MTS, Johnson JA, Balko SU.. Chapter 9 : Diabetes and Eye Disease in Alberta. In: Johnson JA,editor Alberta Diabetes Atlas 2011. Edmonton: Institute of Health Economics 2011. p. 153-74. 10. Bond SR, Advani A, Altomare F, Stockl F. Canadian Diabetes Association 2013 Clinical Practice Guidelines for the Prevention and Management of Diabetes in Canada: Retinopathy. Can J Diabetes. 2013;37(suppl1):S137-S41. 11. Hooper P, Boucher MC, Cruess A, Dawson KG, Delpero W, Greve M, et al. Canadian Ophthalmological Society evidence-based clinical practice guidelines for the management of diabetic retinopathy. Can JOphthalmol. 2012;47(2 Suppl):S1-30, S1-54. 12. Bragge P, Gruen RL, Chau M, Forbes A, Taylor HR. Screening for presence or absence of diabetic retinopathy: a meta-analysis. Arch Ophthalmol. 2011;129(4):435- 44.

 153  

13. Jones S, Edwards RT. Diabetic retinopathy screening: a systematic review of the economic evidence. Diabet Med. 2010;27(3):249-56. 14. McCarty CA, Lloyd-Smith CW, Lee SE, Livingston PM, Stanislavsky YL, Taylor HR. Use of eye care services by people with diabetes: the Melbourne Visual Impairment Project. Brit J Ophthalmol. 1998;82(4):410-4. 15. Browing D. Chapter 14. Diabetic Retinopathy. In: Browing D, editor. Diabetic Retinopathy: Evidence-Based Management: Springer; 2010. p. 369-85. 16. Sara A. Does Equity in Healthcare Use Vary across Canadian Provinces? Healthcare Policy. 2008;3(4):83-99. 17. Ramirez M, Ford ME, Stewart AL, Teresi JA. Measurement issues in health disparities research. Health ServRes. 2005;40(5 Pt 2):1640-57. 18. Shavers VL. Measurement of socioeconomic status in health disparities research. J Natl Med Assoc. 2007;99(9):1013-23. 19. Renahy E, Alvarado-Llano B, Koh M, Quesnel-Vallée A. Income and economic exclusion: do they measure the same concept? IntJ Equity in Health. 2012;11(1):1-11. 20. Pampalon R, Hamel D, Gamache P, Raymond G. A deprivation index for health planning in Canada. Chronic diseases in Canada. 2009;29(4):178-91. 21. Renahy E, Alvarado-Llano B, Koh M, Quesnel-Vallee A. Income and economic exclusion: do they measure the same concept? Int J Equity Health. 2012;11:4. 22. Jotkowitz AB, Rabinowitz G, Segal AR, Weitzman R, Epstein L, Porath A. Do Patients with Diabetes and Low Socioeconomic Status Receive Less Care and Have Worse Outcomes? A National Study. Am J Med. 2006;119(8):665-9. 23. Perry M, Williams RL, Wallerstein N, Waitzkin H. Social Capital and Health Care Experiences Among Low-Income Individuals. Am JPublic Health. 2008;98(2):330-6. 24. van Doorslaer E, Masseria C. Income-Related Inequality in the Use of Medical Care in 21 OECD Countries. Paris:Organisation for Economic Co-operation and Development; 2004. 25. Knowles J, Bales S, Cuong LQ, OanhTT, LuongDH. Health equity in Viet Nam a situational analysis focused on maternal and child mortality. Ha Long:United Nations Children's Fund; 2009. 26. O'Donnell O, van Doorslear E, Wagstaff A, LindelowM. Chapter 8. The Concentration Index. Analyzing Health Equity Using Household Survey Data : A Guide to Techniques and Their Impementation. Washington: The World Bank; 2008. p. 95-108. 27. O'Donnell O, van Doorslear E, Wagstaff A, LindelowM. Chapter 13. Explanning Socioeconomic-Related Health Inequality : Decomposition of the Concentration Index.Analyzing Health Equity Using Household Survey Data : A Guide to

 154  

Techniques and Their Impementation. Washington: The World Bank; 2008. p. 159- 64. 28. Organization for Economic Co-operation and Development. Growing unequal: income distribution and poverty in OECD Nations. Paris: 2008. 29. Haper S, Lynch J. Selected Comparisons of Measures of Health Disparities : A review Using Database Relevant to Healthy People 2010 Cancer-Related Objectives. Bethesda, MD: U.S Deparment of Health and Human Services, National Institutes of Health; 2007. 30. Wagstaff A, Paci P, van Doorslaer E. On the measurement of inequalities in health. Soc Sci Med. 1991;33(5):545-57. 31. Wagstaff A. Socioeconomic inequalities in child mortality: comparisons across nine developing countries. Bulletin of the World Health Organization. 2000;78(1):19. 32. McGrail KM, van Doorslaer E, Ross NA, Sanmartin C. Income-related health inequalities in Canada and the United States: a decomposition analysis. Am JPublic Health. 2009;99(10):1856-63. 33. Jimenez-Rubio D, Smith PC, Van Doorslaer E. Equity in health and health care in a decentralised context: evidence from Canada. Health economics. 2008;17(3):377-92. 34. McGrail KM. Income-related inequities: Cross-sectional analyses of the use of medicare services in British Columbia in 1992 and 2002 Open Medicine 2008;2(4):e91-8. 35. van Doorslaer E, Wagstaff A, Bleichrodt H, Calonge S, Gerdtham U-G, Gerfin M, et al. Income-related inequalities in health: some international comparisons. J Health Econ. 1997;16(1):93-112. 36. Ravaghi V, Quinonez C, Allison PJ. The magnitude of oral health inequalities in Canada: findings of the Canadian health measures survey. Community Dent Oral Epidemiol 2013. 37. Kiran T, Kopp A, Moineddin R, Victor JC, Campbell RJ, Shah BR, et al. Unintended consequences of delisting routine eye exams on retinopathy screening for people with diabetes in Ontario, Canada. CMAJ. 2013;185(3):E167-E73. 38. Oliver A, Mossialos E. Equity of access to health care: outlining the foundations for action. J Epidemiol Community Health. 2004;58(8):655-8. 39. Mooney G. Equity in health care: confronting the confusion. Effective Health Care. 1983;1:179-85.